G1P3, an interferon inducible gene 6-16, is expressed in gastric cancers and inhibits mitochondrial-mediated apoptosis in gastric cancer cell line TMK-1 cell

Abstract

Expression of an interferon inducible gene 6-16, G1P3, increases not only in type I interferon-treated cells but also in human senescent fibroblasts. However, the function of 6-16 protein is unknown. Here we report that 6-16 is 34 kDa glycosylated protein and localized at mitochondria. Interestingly, 6-16 is expressed at high levels in gastric cancer cell lines and tissues. One of exceptional gastric cancer cell line, TMK-1, which do not express detectable 6-16, is sensitive to apoptosis induced by cycloheximide (CHX), 5-fluorouracil (5-FU) and serum-deprivation. Ectopic expression of 6-16 gene restored the induction of apoptosis and inhibited caspase-3 activity in TMK-1 cells. Thus 6-16 protein has anti-apoptotic function through inhibiting caspas-3. This anti-apoptotic function is expressed through inhibition of the depolarization of mitochondrial membrane potential and release of cytochrome c. By two-hybrid screening, we found that 6-16 protein interacts with calcium and integrin binding protein, CIB/KIP/Calmyrin (CIB), which interacts with presenilin 2, a protein involved in Alzheimer’s disease. These protein interactions possibly play a pivotal role in the regulation of apoptosis, for which further detailed analyses are need. These results overall indicate that 6-16 protein may have function as a cell survival protein by inhibiting mitochondrial-mediated apoptosis.

Introduction

Interferon inducible gene 6-16, G1P3, was first identified as one of the genes that are induced by interferon α and β [1]. It locates on chromosome 1p35 [2] and produces three types of mRNA by differential splicing between exon2 and exon3 [3]. The most abundant mRNA (B type) encodes a 14 kDa hydrophobic protein of 130 amino acids. Although extensive studies on the upstream regulatory region of 6-16 gene are well reported [4], little is known that the function or role of 6-16 protein in interferon treated cells. We isolated several cDNA clones that are expressed at higher level in senescent human fibroblasts than in young counterparts and found that one of these cDNA clones is 6-16 [5]. One of the characteristics of senescent human fibroblasts is that the cells do not proliferate but are viable for relatively long period of time, usually for more than a year in a culture dish, without apparent apoptosis. SV40 large T antigen overcomes normal senescent pathway by inactivating some proteins, which are associated with cell proliferation, such as tumor suppressor gene p53 and pRB proteins. Interestingly, 6-16 is still expressed in SV40-transformed cells after extending their life-span [6]. 6-16 expression may be independent from p53 pathway and/or pRB pathway. Interestingly, we found that human senescent fibroblasts express 6-16 by producing interferon-β by autocrine mechanism. Interferon regulatory factor 1, IRF-1, is a major regulator of the interferon signaling pathway. Treatment of anti-IRF-1 antibody to human senescent cells or life-extended SV40-transformed fibroblasts resulted in down-regulation of 6-16 expression [6]. It is still unknown how senescent cells are protected from apoptosis, because senescent cells could be attached to the dish and are viable for more than 1 year with just a medium change; hence, it is believed that some anti-apoptotic gene may be protecting the senescent cells from apoptosis.

Attenuation of apoptosis appears recession for establishment and maintenance of transformed phenotype [7]. We found that 6-16 gene is expressed at a high level in immortalized cells and in gastrointestinal tumor cells. These data suggest that an increase in expression of 6-16 is associated with attenuation of apoptosis.

Mitochondria is an important organelle for the control of apoptosis, in addition to the role as the center of energy metabolism, and influence the commitment of cell death by regulating the mitochondrial permeability and membrane potential [8–11]. Bcl-2, an anti-apoptotic protein, is known to be located on mitochondria and expressed at high level in some tumor cells and tissues [11]. We report here that 6-16 is relatively expressed in cancer cells and tissues, and 6-16 is a novel anti-apoptotic protein located in the mitochondria and can be a new target for cancer chemotherapy and mitochondrial diseases.

Materials and methods

Northern blotting and RT-PCR

RNA isolation and Northern blotting were performed as described previously [6]. All 6-16 splice valiants are recognized by the probe for Northern analysis. RT-PCR was performed according to the protocol with the Thermoscript one-step RT-PCR system (Invitrogen, USA). For the detection of B and C type of 6-16 spliced valiants, we used 5′-GGGTGGAGGCAGGTGAGAATGCGG-3′ as the forward primer and 5′-TGACCTTCATGGCCGTCGGAGGAG-3′ as the reverse primer. Samples were incubated at 50°C for 30 min and denatured at 94°C for 2 min, and then cycled 32 times with 30 s at 94°C and 30 s at 62°C followed by a final extension step of 5 min at 68°C. GAPDH was used for internal control for validating RNA amounts as described before [12].

In situ mRNA hybridization analysis

In situ mRNA hybridization (ISH) was performed as described previously [13] with minor modification. Briefly, an interferon inducible gene 6-16-specific oligonucleotide probe was designed complementary to the 5′-end of human 6-16 mRNA transcript (GenBank NM022873). The DNA oligonucleotide sequence 5′-CGCCGCCCCCATTCAGGA-3′ was of the antisense orientation and hence complementary to 6-16 mRNA. To verify the integrity and lack of degradation of mRNA in each sample, we used a d(T) oligonucleotide. All DNA probes were synthesized with six biotin molecules (hyperbiotinylated) at the 3′-end via direct coupling using standard phosphoramidite chemistry (Research Genetics, Huntsville). ISH was carried out using the Microprobe manual staining system (Fisher Scientific, Pittsburgh). A positive reaction in this assay stained red. Control for endogenous alkaline phosphatase included treatment of sample in the absence of the biotinylated probe and use of chromogen alone. To check the specificity of the hybridization signal, the following controls were used: RNAse pretreatment of tissue sections, substitution of the antisense probe with a biotin-labeled sense probe, and competition assay with unlabeled antisense probes. No or markedly decreased signals were obtained after either of these treatments.

Antibody

Rabbit polyclonal antibody against human 6-16 was raised against synthetic peptide (YATHKYLDSEEDEE) corresponding to amino acid residues 117–130 of human 6-16, and was purified by MAbTrap GII affinity chromatography kit (Amersham Bioscience, USA). This antibody was sufficient for immunoblotting but insufficient for immunoprecipitation and immunostaining of 6-16 protein. Another rabbit polyclonal anti-human 6-16 antibody (OT904-1A) was also prepared against synthesis peptides (VEAGKKKCSESSDSG) corresponding to amino acid residues 21–35 of human 6-16. This antibody was sufficient for immunostaining. Mouse monoclonal anti-human CIB antibody was kindly provided by Leslie V. Parise. Anti-cytochrome c antibody (clone 7H8.2C12 which recognizes the denatured form of human, mouse and rat cytochrome c; BD Pharmingen, USA), monoclonal anti-human Bcl-2 antibody (clone124, Upstate Biotechnology, USA) and polyclonal anti-Bax antibody (Upstate Biotechnology) were used for immunoblot and immunoprecipitation. Anti-cytochrome c antibody (clone 6H2.B4 which recognizes the native form of human, mouse and rat cytochrome c; BD Pharmingen) was used for immunostaining. Polyclonal anti-GST rabbit antibody (New England Biolabs, USA) and polyclonal anti-MBP antibody (kindly provided by Dr. M. Nakata, Sumitomo Electric Industries, Japan) were used for immunoblot.

Cell culture, PI staining, caspase-3 activity assay, and mitochondrial membrane potential assay

The 0.8 kb fragment encoding full length 6-16 was inserted in the pCXN vector, which was digested with Xho I and blunted with klenow fragment. To establish stable transfected TMK-1 cells expressing 6-16 (TMK-1-6-16), TMK-1 cells were transfected with pCXN/6-16 vector and were selected for neomycin resistance with 200 μg/ml Geneticin (Invitrogen, USA). Gastric cancer cell lines, TMK-1, TMK-1-6-16 and MKN-28 were cultured at 37°C and 5% CO₂ in RPMI 1640 (Invitrogen) supplemented with 10% FCS and Antibiotic-Antimycotic (Invitrogen). A normal human fetal fibroblast strain, TIG-3, was cultured in Dulbecco’s modified Eagle medium (Invitrogen) supplemented with 10% FCS and Antibiotic-Antimycotic (Invitrogen).

Apoptosis was induced by actinomycin D (10, 20 and 40 μg/ml) for 24 h, cycloheximide (10, 30 and 100 μM) for 6 h, H₂O₂ (0.5, 5 and 50 μM) for 6 h, etoposide (200, 400 and 600 μM) for 24 h, bleomycin (100, 200 and 400 μg/ml) for 24 h, 5-FU (100, 200 and 400 μg/ml) for 24 h, aphidicolin (100, 200 and 400 μM) for 24 h or serum-deprivation for 60 h. Cells were collected and stained with PI-RNase solution (BD Biosciences, USA), and were analyzed for DNA content by Flow cytometry on a FACS Calibur. For DNA ladder analysis, apoptotic DNA was extracted by lysis buffer (50 mM Tris-HCl pH7.5/ 20 mM EDTA/1% NP40), and run on 2% agarose gel.

For Caspase-3 activity analysis, cells were grown in six-well plates overnight, and were treated with CHX (10 and 30 μM) for 6 h or with 5-FU (80 μg/ml) for 24 h. Cells were collected and incubated with PhiPhiLux substrate buffer (OncoImmunin Inc., Gaithersbury, MD, USA) for 60 min at 37°C in 5% CO₂ incubator, and were analyzed by Flow cytometry on a FACS Calibur.

For mitochondrial membrane potential analysis, cells were grown in six well plate overnight, and were treated with CHX (10 and 30 μM) or 5-FU (50 and 100 μg/ml) for an hour. Cells were collected and incubated with Mitosensor reagent buffer (Clontech) for 30 min at 37°C in CO₂ incubator, and were analysis by Flow cytometry on a FACS Calibur.

Immunofluorescence analysis

Cells grown on eight-well Lab-Tek chamber slides (Nalge Nunc International, Napaperville, IL) were fixed in 4% paraformaldehyde / PBS (pH 7.4) for 20 min and permeabilized in blocking buffer (0.2% Triton X-100/3% BSA/PBS) for 30 min. The cells were incubated for 1 h at room temperature with mouse anti cytochrome c monoclonal antibody (6H2.B4) diluted 1:100, or with rabbit anti-human 6-16 polyclonal antibody (OT904-1A) diluted 1:100 in blocking buffer. The cells were washed four times in PBS and incubated with Alexa FluorTM 488 goat anti-mouse IgG conjugate (Molecular Probes, USA) or with Alexa FluorTM 488 goat anti-rabbit IgG conjugate (Molecular Probes) diluted 1:200 in blocking buffer for an hour at room temperature. Immunofluorescent staining was analyzed using a confocal laser-scanning microscope (LSM510, Carl Zeiss, Germany).

Mitochondria were labeled in intact cells with MitoTracker CM-H₂XRos (Molecular Probes). Cells were incubated in the MitoTracker medium (final concentration 500 nM) for 45 min before finishing to treat with CHX or 5-FU. Then, the cells were fixed, permeabilized and double-labeled with anti-cytochrome c antibody (6H2.B4) as described above.

Cell fractionation, immunoprecipitation and immunoblot analysis

For isolation of cellular fraction, cells were suspended in sucrose-supplemented extraction buffer (SCEB, 300 mM sucrose, 10 mM HEPES pH 7.4, 50 mM KCl, 5 mM EGTA, 5 mM MgCl₂, 1 mM DTT, protease cocktail), left on ice for 30 min, and homogenized by 70 strokes in an ice-cold Dounce homogenizer. Unbroken cells and nuclei were pelleted by centrifugation for 10 min at 2,000×g, and the supernatant was further spun at 13,000×g for 20 min to separate mitochondria-rich fraction from cytosol fraction. The pellet was resuspended in 50 μl of SCEB (mitochondria fraction), and the supernatant was cytosolic fraction.

For co-immunoprecipitation experiments, cells were lysed in NP-40 lysis buffer (50 mM Tris-HCl, pH7.4, 120 mM NaCl, 1% NP-40) containing protease inhibitor cocktail (Roche, USA). Antigen-antibody reaction was performed by incubating 0.5 ml (500 μg protein) of the cell extract with rabbit anti-human 6-16 antibody (117–130), mouse anti-CIB antibody, rabbit anti-Bax antibody or mouse anti-Bcl-2 antibody overnight at 4°C. The immunocomplex were incubated with 50 μl protein-G-Sepharose for 3 h at 4°C, and the beads were washed three times with NP-40 lysis buffer and boiled in Laemmli buffer.

For immunoblot analysis, the proteins were separated on 13% or 15% SDS-PAGE, and transferred to Immobilon-P (Millipore) and immunoblotted with anti-6-16 antibody (117–130), anti-CIB antibody, anti-Bcl-2 antibody, anti-Bax antibody or anti-cytochrome c antibody (7H8.2C12) diluted 1:500. The signal was detected using ECL-Plus (Amersham Bioscience).

Yeast two-hybrid assay

Dr Y. Takai (Osaka University, Suita, Japan) kindly supplied yeast L-40 strain and pBTM116/HA for yeast two-hybrid screening. A strain of L40 carrying pBTM116/6-16 was transformed with pGADGH HeLa cDNA library (Clontech). Approximately 1×10⁶ transformants were screened for the growth on SD medium plated lacking Trp, Leu, and His as evidenced by transactivation of a LexA-HIS3 reporter gene and His prototrophy. His+ colonies were scored for β-galactosidase activity. Plasmids harboring cDNAs were recovered from positive colonies and introduced by electroporation into E. coli HB101 on the M9 plate lacking Leu. Then the plasmids were recovered from HB101 and transformed again into L40 containing pBTM116HA/6-16. The nucleotide sequences of plasmid DNAs were determined.

Interaction of proteins in vitro

6-16-pGEX-2T (Amersham Bioscience) or CIB-pMAL-C2 (New England Biolabs) expression vectors were constructed to produce GST-6-16 and MBP-CIB proteins. GST-6-16 and MBP-CIB proteins were purified from E. coli transformed with their expression vectors treated with 0.1 M IPTG by using amylose resin (New England Biolabs) or glutatione Sepharose 4B (Amersham Bioscience). GST-6-16 protein (20 pmol) was incubated with MBP-CIB protein (40 pmol) in 40 μl of reaction buffer (20 mM Tris-HCl pH7.5, 1 mM DTT and 0.05% CHAPS) for 1 h at 4°C. After glutathione Sepharose 4B (Amersham Bioscience) was added and further incubated for 1 h, the precipitates were washed three times and subjected to immunoblot analysis using anti-MBP and anti-GST rabbit polyclonal antibody.

Results

6-16 was expressed in gastric cancer cells and tissues

We examined expression levels of 6-16 mRNA in eight gastric cancer cell lines (MKN-1, MKN-7, MKN-28, MKN-45, MKN-74, TMK-1, HSC-39 and KATO III) by Northern blot analysis. Seven out of eight gastric carcinoma cell lines expressed 6-16 mRNA, while TMK-1 cells showed very low levels (Fig. 1a). p53 mutation (codon 173, GTG to ATG) was found in TMK-1 cells [14]. Other gastric cell lines also have p53 mutation excluding MKN-45 [14]. However, there is no correlation between p53 mutation status and 6-16 expression levels. TMK-1 cell is sensitive to apoptosis compared with other gastric cancer cell line used in Fig. 1a. In the other cancer cell lines that were frequently used for apoptosis research, colorectal cancer cell line RKO, which is sensitive to apoptosis, express low levels of 6-16. In contrast, high levels of 6-16 expression was found in breast cancer cell line MCF-7 cell, which is resistant to apoptosis (Fig. 1b). Interestingly, HL-60 cell line does not have a significant level of 6-16 expression, but anti-cancer drug resistant clone of HL-60 cells do have high levels of 6-16 expression (Fig. 1b). Human uterine sarcoma cell line, NES-SA and the multiple drug-resistant uterine sarcoma cell line, MES-SA/DX7 also express high levels of 6-16, but there is significant difference between these two cell lines. Taken together, there is good correlation between 6-16 expression levels and resistant to apoptosis. We next studied the expression of 6-16 mRNA in primary gastric carcinomas by in situ mRNA hybridization (ISH) (Fig. 1c). In almost all of these tumors, higher expression of 6-16 mRNA was shown in cytoplasm of the tumor cells in comparison with the corresponding non-neoplastic mucosas. We also recognized weak signal in stromal fibroblast cells and fundic gland cells, but not in muscular tissues of the gastrointestinal tracts (Fig. 1c).

Fig. 1.

Expression of 6-16 mRNA in gastric cancer cells and carcinoma tissues. a Expression of 6-16 mRNA was detected by Northern blot analysis with 10 μg total RNA in eight gastric cancer cell lines. Hybridization of a G3PDH (control) probe to the same filter membrane is depicted in the lower panel. b Expression of 6-16 mRNA was detected by RT-PCR analysis. 0.2 μg of RNA were used for RT-PCR reaction. PCR products were run on 6% acrylamide gel electrophoresis and staining with CYBR Green I nucleic acid staining. c Surgically resected adenocarcinomas of the stomach were examined on in situ hybridization using the microprobe manual staining system (Fisher Scientific). 6-16-specific anti-sense oligonucleotide DNA probe (5′-GCA CGC CGC CCC CAT TCA GGA TCG CAG-3′) was designed. lower panel 6-16 in situ hybridization. upper panel hematoxylin-eosin (HE) staining

6-16 is 32 kDa glycosylated protein

To elucidate the function of 6-16, we established 6-16 expressing TMK-1 cells by transfection of 6-16 cDNA. After selection with G418, we isolated several clones and examined the expression of 6-16. Among them, clone 8 expresses the highest level of 6-16 mRNA (Fig. 2a). Isolated clones in order of 6-16 expression levels are No. 8, 11 and 3. Clone 11 also expresses 6-16, but slightly lower than clone 8. Expression of 6-16 in clone 3 is lowest in these clones. We used these three types of clones to examine the function of 6-16. Unless otherwise indicated, TMK-1-6-16 clone 8 is referred to as TMK-1-6-16, and is used for further examination in comparison with the parental TMK-1.

Fig. 2.

Localization of 6-16 at mitochondria. a Total RNA was isolated from each cell line, and 10 μg of each was used for Northern blot analysis. Semi-quantitative analysis of 6-16 mRNA level of autoradiographs was performed using the public domain NIH Image program. The units are arbitrary, and were calculated based on the expression of 6-16 mRNA in TMK-1 cells as 1.0. Hybridization of a G3PDH (control) probe to the same filter membrane is depicted in the lower panel. b In vitro transcription/translation products were synthesized from luciferase control (positive control: P.C), pZero/Kan (negative control: N.C) and 6-16-pZero/Kan expression vector by using TNT coupled Wheat Germ Extract System (Promega). Incorporation products of ³⁵S-methionine were detected by autoradiography after separating by 15% SDS-PAGE (left panel). Non-radioactive products were detected by immunoblot analysis with anti-6-16 antibody (right panel). C, Extracts from TIG-3, TMK-1, TMK-1-6-16 and MKN-28 cells were incubated with different concentrations of glycosidase at 37°C for 48 h, and detected by immunoblot analysis with anti-6-16 antibody

To analyze 6-16 protein expression, we generated 6-16 polyclonal antibody that was raised against synthetic peptide (YATHKYLDSEEDEE) corresponding to amino acid residues 117–130 of human 6-16 and was purified by MAbTrap GII affinity chromatography. Theoretical molecular weight of 6-16 is about 14 kDa protein by using ExPASy molecular biology database server. By using in vitro transcription/translation experiments from 6-16 cDNA, 14 kDa protein was detected both by autoradiography of ³⁵S-methionine-incorporated products and immunoblotting using a specific antibody (data not shown). By using cell extract from human culture cell, we could not detect 14 kDa 6-16 protein by Western blotting analysis, but 32 kDa protein was detected in 6-16 highly expressing cells such as TMK-1-6-16 and MKN-28. TIG-3 and TMK-1 cells are low or no 6-16 expression by immunoblotting analysis (Fig. 2b). Thus, it is speculated that 6-16 has some protein modifications such as phosphorylation or glycosilation. By the ProDom database analysis (motif), 6-16 protein has multiple possible glycosylation sites including 15 serine residues and 5 threonine residues. To clarify this possibility that 6-16 protein is glycosilated, we treated glycosidase to the cell lysates that were isolated from TMK-1 and TNK-1-6-16 cells, and performed SDS-PAGE and immunoblotting. By digestion with glycosidase, the shift of 32 kDa protein band to 14 kDa band was observed with increasing enzyme concentration, indicating an extensive glycosylation of the native 6-16 protein (32 kDa) in cells (Fig. 2b). These results suggest that 6-16 protein is 34 kDa glycosilated protein.

Apoptosis was attenuated in cells expressing 6-16 protein at high level

6-16 protein is expressed at high levels in both tumor cells and senescent cells. As shown in Fig.1, most of the gastric cancer cell lines expressed 6-16 at high levels except TMK-1 cell. So, we examined whether overexpression of 6-16 protein has an effect on the apoptosis. Out of eight inducers of apoptosis field such as actinomycin D, cycloheximide (CHX), H₂O₂, etoposide, bleomycin, 5-fluoro-2′-deoxyuridine(5-FU), aphidicolin or serum-deprivation, only CHX, 5-FU and serum-deprivation induced apoptosis in TMK-1. However, in TMK-1-6-16 cells which are overexpressing 6-16 protein, apoptosis induced by CHX or 5-FU were significantly inhibited by subG0 analysis using Flow cytometry (Fig. 3a). The other five inducers of apoptosis did not induce apoptosis in either TMK-1 and TMK-1-6-16 cells (data not shown). The increase in subG0 fraction was observed in the dose-dependent manner after CHX or 5-FU treatment in TMK-1, but not in TMK-1-6-16 cells (Fig. 3a). TMK-1-6-16 (clone 3) and TMK-1-6-16 (clone 11) expressed less 6-16 (Fig. 1b) and was less resistant to CHX or 5-FU induced apoptosis than TMK-1-6-16 (Fig. 3a). MCF-7 expressed 6-16 at high levels and was well known to apoptosis-resistant cancer cell line (Fig. 1b). In contrast to MCF-7, RKO that is well known to apoptosis sensitive cancer cell line expressed 6-16 at low levels (Fig. 1b). In addition, TMK-1 is most sensitive cancer cell line in gastric cancer cell line which showed in Fig. 1a (data not shown). Taken together, there is good correlation between 6-16 expression levels and resistance of apoptosis. The amount of DNA extracted in NP-40 buffer, called DNA ladder, was increased in TMK-1 cells after CHX or 5-FU treatment, but not in TMK-1-6-16 (Fig. 3b). The extracted DNA showed a characteristic ladder found in DNA from apoptotic cells (Fig. 3b). DNA ladder formation is dose dependent manner (data not shown). These results indicate that 6-16 protein attenuates apoptosis induced by CHX or 5-FU in TMK-1 cell.

Fig. 3.

6-16 inhibits apoptosis and caspase-3 activity by CHX and 5-FU. a TMK-1 and TMK-1-6-16 cells were treated with CHX for 6 h or with 5-FU for 24 h. Cells were collected and stained with PI-RNase solution, and were analyzed for DNA content by Flow cytometry. Percentage of cells with subG0 DNA content was calculated. b TMK-1 (T) and TMK1-6-16 (6) cells were treated with CHX (10 and 30 μM) for 6 h or with 5-FU (100 and 200 μg/ml). NP-40 (1%) extracted DNA was purified and run on 2% agarose gel. M indicates 100 bp DNA ladder of size marker. c TMK-1 and TMK-1-6-16 cells were treated with CHX (10 and 30 μM) for 6 h or with 5-FU (80 μg/ml) for 24 h. Cells were collected and were measured for caspase-3 activity by PhiPhilux G1D2 Kit by using Flow cytometry. Cells with FL-1H amp gain over 180 were referred to as caspase-3 positive cells

6-16 localized at mitochondria and co-localized with cytochrome-c

The PSORT II and TMpred database analysis have shown that 6-16 protein has transmembrane helices, a mitochondria targeting sequence and an intramitochondrial sorting signal(Fig. 4a). It was quite interesting that, from database analysis on amino acid sequence, 6-16 protein has four transmembrane domains, a mitochondrial targeting sequence and an intra-mitochondrial sorting signal (Fig. 4a). So, it is possible that 6-16 is localized at mitochondria. To check this possibility, we have examined subcellular localization of 6-16. These data led us to examine intracellular localization and possible role on apoptosis of 6-16 protein. By immunofluorescence, 6-16 protein was co-localized with cytochrome c in both TMK-1 and TMK1-6-16 cells (Fig. 4b). Essentially, the same results were obtained using a colon cancer cell line, RKO, and a breast cancer cell line, MCF-7 (data not shown). 6-16 protein was not detected in normal human fibroblast, TIG-3, at 32 PDL (Fig. 4b) but was weakly detected at 75 PDL (data not shown), consistent with the previous results that the expression of 6-16 mRNA increased with cellular senescence [6]. Mitochondrial localization of 6-16 protein, as well as cytochrome c, was confirmed by immunoblotting of subcellular fractionations (Fig. 4c). Taken together with these data, 6-16 is mainly localized in mitochondria and co-localized with cytochrome c.

Fig. 4.

6-16 co-localized with cytochrome c at mitochondria. a 6-16 protein has possible glycosylation sites (15 serine residues and 5 threonine residues) and possible mitochondria localization site [one intramitochondrial target sequence, one APOLAR (apolar signal of intramitochondria sorting) signal domain and four mitochondrial helices domains]. Open circle serine residues, square threonine residues, black bar transmembrane helices domain, gray bar intramitochondrial target sequence, dotted bar APOLAR signal domain. b TIG-3 (32PDLs), TMK-1 and TMK-1-6-16 cells were fixed with 4% paraformaldehyde, and permeabilized in 0.3% Triton-X100. 6-16 (red) and cytochrome c (green) were detected by immunofluorescence. Co-localization was seen by merged image of the green and red signals (yellow). c Localization of 6-16 protein was detected by immunoblot analysis of whole-cell lysates (total), mitochondria fraction (M) and cytosolic fraction (C). All fractions were adjusted to 20 μg proteins and analyzed by immunoblotting with anti-6-16 antibody or anti-cytochrome c antibody

6-16 protein inhibits cytochrome c release and reduction of mitochondrial membrane potential

To search for the mechanism of anti-apoptotic function of 6-16, we first examined caspase-3 activity. The percentage of cells positive for caspase-3 activity determined by Flow cytometry was remarkably increased in TMK-1 cells after the treatment with CHX and 5-FU (Fig. 3c). The increase was seen in TMK-1-6-16 only at high concentration of CHX at 30 μg/ml. These results suggested that 6-16 inhibited the caspase-3 dependent apoptotic pathway.

The apoptosis related-proteins (e.g. cytochrome c, caspase-2, −3, −9, Hsp10, Smac/DIABLO and AIF) are released from mitochondria into the cytosol during apoptosis [10, 15], and the activation of the caspase cascade is dependent on the release of cytochrome c from mitochondria [10, 16, 17]. Immunoblot analysis demonstrated that in TMK-1 cells, cytochrome c in mitochondria was released into cytosolic after treatment with CHX or 5-FU (Fig. 5a). In TMK-1-6-16 cells, however, cytochrome c was detected in mitochondria after either treatment (Fig. 5a). 6-16 protein remained in mitochondria fraction in both TMK-1 and TMK-1-6-16 cells after treatment with either CHX or 5-FU (Fig. 5a). Release of cytochrome c was also monitored by immunofluorescence staining. In non-apoptotic TMK-1 cells, the staining patterns of mitochondria (red) and cytochrome c (green) were completely overlapped (merge: yellow) (Fig. 5b). After treatment with CHX or 5-FU, the signals of cytochrome c were distinct from that of mitochondria in TMK-1 cells (Fig. 5b left panel), whereas these two signals always overlapped in TMK-1-6-16 cells (Fig. 5b right panel). When TMK-1 cells were treated with CHX or 5-FU for an hour, cytochrome c was diffusely observed in cytosol. After 3 h, cytochrome c signal appeared in granules or grains in cytosol distinct from mitochondria, indicating that cytochrome c in cytosol might interact with the protein complex including Apaf-1/pro-caspase-9 protein [16].

Fig. 5.

6-16 protein inhibits release of cytochrome c and reduction of mitochondrial membrane potential. a For immunofluorescence, cells were treated with CHX (10 μM) and 5-FU (200 μg/ml), labeled with MitoTracker CM-H₂XRos (Red) for 45 min, and fixed with 4% paraformaldehyde, permeabilized in 0.3% Triton-X100, and stained with anti-cytochrome c antibody (green). b TMK-1 and TMK-1-6-16 cells were treated with CHX (10 μM) for 6 h or with 5-FU (80 μg/ml) for 24 h. Cells were collected and separated to mitochondria (M) and cytosolic (C) fractions. Twenty micrograms of protein was used for immunoblot analysis with anti-cytochrome c antibody or anti-6-16 antibody. c TMK-1 and TMK-1-6-16 cells were treated with CHX (10 and 30 μM), 5-FU (50 and 100 μg/ml) for an hour. Cells were collected and incubated with Mitosensor reagent buffer (Clontech) for 30 min at 37°C in CO₂ incubator, and Δ Ψ m was analyzed by Flow cytometry (left panel). Cells with fluorescence intensity below 30 were referred to as FL-2H negative cells (right panel). Error bars represent standard deviations from three independent samples

As the release of cytochrome c from mitochondria is well associated with depolarization of mitochondrial membrane potential (Δ Ψ m) [8–10], we next measured mitochondrial membrane potential at single cell level by Flow cytometry after staining cells with JC-1 or MitoSensor reagent [18, 19]. In TMK-1 cells, a cell population with decreased Δ Ψ m (depolarization) appeared after the treatment with CHX or 5-FU, but not in TMK-1-6-16 cells (Fig. 5c). These results overall suggested that 6-16 protein inhibits cytochrome c release from mitochondria by inhibiting depolarization of mitochondrial membrane potential (Δ Ψ m), resulting in attenuation of apoptosis.

6-16 interacts with calcium integrin binding protein, CIB

In order to further understand the anti-apoptotic mechanism of 6-16 protein, we screened 6-16 interacting protein by yeast two-hybrid methods. The fusion protein of full length 6-16 and GAL4 (GAL4-6-16) was used as bait. By the screening of 1×10⁶ colonies of HeLa cDNA library, 41 positive clones (His+ and LacZ+) were identified. Among them, twenty-seven cDNA clones encoded CIB/KIP/calmyrin (we refer to it as CIB in this paper) (Genbank: U822226, U85611). CIB (calcium and integrin binding protein) [20] was also reported from independent groups as DNA-PK interaction protein (KIP) [21] and calcium-binding myristoylated protein with homology to calcineurin (calmyrin) [22]. Twelve cDNA clones encoded γ-subunit of the eukaryotic cytosolic chaperonin-containing protein, TCP-1 (CCTγ) (EMBL:X74801), one cDNA clone encoded calcium binding protein related S-100 (CAPL) protein (GenBank : M80563) and one was an unknown gene.

We next examined the association of 6-16 and CIB proteins in vitro. For this purpose, we prepared GST-tagged 6-16 protein (GST-6-16) and MBP-tagged CIB protein (MBP-CIB) by purification from E. coli transformed with each construct and assayed for in vitro binding. As shown in Fig. 6a, GST-6-16 and MBP-CIB proteins co-immunoprecipitated in vitro, and 6-16 interacted with CIB using by immunoprecipitation with MKN-28 cell extracts in vivo (Fig. 6b).

Fig. 6.

CIB protein binds to 6-16 and Bcl-2 protein. a GST-6-16 and MBP-CIB proteins were purified from E. coli, and were incubated in vitro, precipitated with glutathione sepharose 4B and detected by immunoblot analysis with anti-GST antibody (lower panel) or anti-MBP antibody (upper panel). b MKN-28 cell extracts were immunoprecipitated with antibody against CIB, Bcl-2 or Bax (top margin of the panel), and subjected to immunoblot analysis with anti-6-16 antibody, anti-Bcl-2 antibody or anti-Bax antibody (right margin of the panel)

We demonstrated that 6-16 protein was localized at mitochondria and attenuated apoptosis by controlling Δ Ψ m and cytochrome c release. The Bcl-2 family proteins are known to control Δ Ψ μ and cytochrome c release via VDAC. We then hypothesized that 6-16 and/or CIB proteins might interact with Bcl-2 family proteins. For this purpose, we searched for cell lines that expressed both 6-16 and CIB at high level and found a gastric cancer cell line MKN-28, which also expressed Bcl-2 at high level (Fig. 5b). By immunoprecipitation with anti-CIB antibody, anti-Bcl-2 antibody or anti-Bax antibody followed by immunoblotting, we found that 6-16 protein interacted with CIB, but not with Bcl-2 or Bax (Fig. 6b). CIB protein interacted with 6-16 and weakly with Bcl-2, but not with Bax (Fig. 6b). Bcl-2 interacted weakly with CIB and Bax (Fig. 6b).

Discussion

In this study, we found that 6-16 protein inhibits apoptosis induced by 5-FU or CHX in gastric cancer cell line TMK-1 cells through the mitochondrial pathway. There are four different inhibition categories of apoptosis. First, in the death receptor pathway (e.g. Fas, TNF and TRAIL), RIP (receptor-interacting protein), c-FLIP (cellular-Flice-like inhibitory protein) and FAP-1 (Fas-associated phosphatese-1) inhibit death signal from the death receptors nearly located at the plasma membrane [23, 24]. Second, in the mitochondrial pathway, anti-apoptotic Bcl-2 family proteins prevent mitochondrial membrane permeabilization to inhibit mitochondrial membrane potential change and cytochrome c release [25]. Third, IAPs (inhibition of apoptosis proteins), including XIAP, cIAP-1 and cIAP-2, selectively inhibit the activity and activation of various caspases [17, 24]. Heat shock proteins (e.g. Hsp10, 27, 60, 70 and 90) also can promote or inhibit caspase activation by altering the conformation of various proteins. Finally, in degradation of chromosomal DNA during apoptosis, ICAD inhibit CAD activity as CAD/ICAD complex [26, 27]. Among these four categories, we found that 6-16 protein inhibits mitochondria-mediated apoptosis as well as anti-apoptotic Bcl-2 family proteins. BH4 domain of anti-apoptotic Bcl-2 family members closes VDAC and inhibits apoptotic mitochondrial changes and cell death [28, 29]. However, 6-16 protein does not have a BH-4 domain by protein domain homology analysis, suggesting that it may function via novel anti-apoptotic mechanisms different from Bcl-2 family proteins. It is important to understand the mechanism of up-regulation of 6-16 expression in cancer cells. Previously, we have found that 6-16 expression was up-regulated in senescent cells through beta interferon signaling pathway, because 6-16 expression was blocked by anti-INF-beta treatment to cultured senescent cells. It is possible that cancer cells may produce beta interferon and induce 6-16 expression by autocrine mechanism, but further examination are need to conclude this possibility.

Cancer and senescence may be viewed as a balance between proliferation and cell death. Many anticancer drugs are designed to induce apoptosis via cytochrome c/Apaf-1/caspase-9 (apoptosome) pathway [16, 24], and the mitochondria play a crucial role for the regulation of tumorigenesis and senescence. Apoptosis induced by CHX is mediated by the apoptosome pathway, whereas 5-FU induced apoptosis is owing to both the Fas/FasL pathway and the apoptosome pathway [24]. 5-FU treatment results in a p53-dependent increase in expression of FasL in human colon cancer cell lines, and apoptosis occurs through Fas/FasL pathway. One of the Fas/FasL pathways also acts to mitochondria after mediating BID and caspase-8. As Fas is expressed in TMK-1 cells (unpublished data), 6-16 protein also inhibits at the mitochondria the apoptosis signal via both the Fas/FasL pathway and the apoptosome pathway in gastric cancer cell line, TMK-1 cells. Therefore, because 6-16 is strongly expressed in almost all the gastrointestinal cancer cells, it is a possibly that 6-16 protein at mitochondria is involved in resistance to anti-cancer drugs. Free radicals, apoptosis inducers, are known to induce cellular senescence and increase with cellular senescence. 6-16 expression increased with senescence and may also inhibit apoptosis by free radical and maintain cell viability in senescence cells.

Interferon (IFN) has antiviral activity and is also known to induce apoptosis. When cells are infected with virus, 6-16 is expressed after induction of IFN. Then, an increase in 6-16 expression may cause resistance against apoptosis after viral infection. If so, IFN has apparently conflicting dual functions of both induction and inhibition of apoptosis.

We identified CIB as a 6-16 interacting protein through yeast two-hybrid screening methods. CIB protein was reported to interact with cytoplasmic domain of integrin αIIbß3 [20], eukaryotic DNA-dependent protein kinase DNA-Pkcs [21], presenilin 2 (PS2) [22], and the polo-like protein kinases Fnk and Snk [30]. The structural properties of CIB indicate that it is a hydrophilic calcium-binding protein with two EF-hand motifs corresponding to the two C-terminal Ca²⁺ binding domains, most similar to calcineurin B (58% similarity) and calmodulin (56% similarity) [20]. Calcineurin is found to dephosphorylate BAD, a pro-apoptotic member of the Bcl-2 family, thus enhancing BAD heterodimerization with Bcl-X_L and promoting apoptosis. Therefore, CIB might possess protein phosphatase activity like calcineurin or other BH3 only group of Bcl-2 family proteins to promote apoptosis. It is of interest that CIB interacts with 6-16, but further examinations are needed to conclude the involvement in the anti-apoptotic activity.

Intracellular Ca²⁺ concentration changes are important as early events in apoptosis, and maintenance of both mitochondrial and ER Ca²⁺ pool is necessary for cell survival [31, 32]. CIB (calmyrin) was found to form the complexes including presenilin, ß- and delta-catenin, p0071, amyloid ß-protein precursor, filamin/Fh-1, Notch, GSK3ß, Rab11, QM/Jif-1 and Bcl-X_L. Moreover, Presenilin 2 interacts with Sorcin, which is a penta-EF-hand Ca²⁺ -binding protein that modulates the ryanodine receptor (RyR) intracellular channel [33]. Presenilin 1 and 2 are well known for their role in Alzheimer’s disease, which is associated with accumulation of ß-amyloid (amyloidogenic Aß42 peptide), abnormality of the mechanism in ER (endoplasmic reticulum) and increased rate of mitochondria-mediated apoptosis in selected areas of the brain [34]. CIB interacted with PS2, and overexpression of CIB and/or PS2 promotes cell death in vitro [22]. In fact, transfection of CIB into TMK-1 and MKN-28 cells also induced apoptosis (unpublished data). 6-16 and CIB may regulate not only mitochondria channels but also Ca²⁺ channels in ER (endoplasmic reticulum). CIB co-localizes and interacts with PS2 localized in the ER [22], and 6-16 possibly localizes at ER membrane because 6-16 has ER membrane retention signal of XXRR-like motif in the N-terminus revealed by PSORT search. Bcl-2 protein is also reported to localize not only at mitochondrial membrane but also at ER and nuclear membrane, and Bcl-2 modulates both mitochondrial and ER Ca²⁺ concentration [31, 32].

Our results indicate that 6-16 and CIB may play a critical role in the regulation of apoptosis via controlling mitochondrial and ER channels through the interaction with Bcl-2 family proteins. Therefore, 6-16 may function as a cell survival protein and CIB as cell death protein. Further examination is needed for understanding the function of interaction of 6-16 protein and CIB protein. However, these data provide a new mechanism in protecting apoptosis. It is possible that the IFN-inducible gene, 6-16 is a new target for cancer therapy and mitochondrial diseases.