Biochimica et Biophysica Acta 1823 (2012) 876–888

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

journal homepage: www.elsevier.com/locate/bbamcr

Role of BI-1 (TEGT)-mediated ERK1/2 activation in mitochondria-mediated and splenomegaly in BI-1 transgenic mice

Jung-Hyun Kim a, Eung-Ryoung Lee a, Kilsoo Jeon a, Hye Yeon Choi a, Hyejin Lim a, Su-Jeong Kim a, Han-Jung Chae b, Seung Hwa Park c, SangUk Kim d, Young Rok Seo e, Jin-Hoi Kim a, Ssang-Goo Cho a,⁎ a Department of Animal Biotechnology, Animal Resources Research Center, and SMART-IABS, Konkuk University, Seoul, South Korea b Department of Pharmacology and Research, Chonbuk National University, Jeonju, South Korea c Department of Anatomy, College of medicine, Konkuk University, Seoul, South Korea d School of Interdisciplinary Bioscience and Bioengineering, Pohang University of Science and Technology, Pohang, South Korea e Department of Life Science and Institute of Environmental Medicine, Dongguk University, Seoul, South Korea article info abstract

Article history: Bax Inhibitor-1 (BI-1) is an evolutionally conserved apoptotic suppressor and belongs to the BI-1 family Received 16 September 2011 of , which contain BI-1-like transmembrane domains. As their cellular functions and regulatory Received in revised form 22 January 2012 mechanisms remain incompletely understood, we compared their anti-apoptotic properties. Forced expression Accepted 23 January 2012 of BI-1 resulted in the most effective suppression of stress-induced apoptosis, compared with other family Available online 28 January 2012 members, together with significant extracellular signal-regulated kinase (ERK)1/2 activation. BI-1-mediated ERK1/2 activation led to the suppression of mitochondria-mediated (ROS) production. Keywords: fi Bax inhibitor-1 Involvement of the ERK signaling pathway in BI-1-induced anti-apoptotic effects was con rmed by knockdown BI-1 family studies with ERK- or BI-1-specific siRNA. Moreover, we produced transgenic (TG) mice overexpressing BI-1, and ERK1/2 the relationship between ERK1/2 activation and the suppression of ROS production or apoptosis was confirmed Splenomegaly in mouse embryonic fibroblast (MEF) cells derived from these mice. Interestingly, we found that BI-1 TG mice Autoimmune response showed splenomegaly and abnormal megakaryopoiesis. Taken together, our results suggest that BI-1-induced ERK1/2 activation plays an important role in the modulation of intracellular ROS generation and apoptotic cell death and may also affect autoimmune response. © 2012 Elsevier B.V. All rights reserved.

1. Introduction superfamily member 6 (Fas) ligation [4].MAPKscontainp42/p44 extracellular signal-related kinase (ERK), c-Jun N-terminal protein Apoptosis is a necessary process for the control of cell growth kinase (JNK)/stress-activated protein kinase (SAPK), and p38 MAPK. and elimination of damaged or unwanted cells, and its dysregulation Several isoforms of MAPK have been characterized in mammals: has been implicated in many diseases, including neurodegenerative ERK1/2, 3/4, 5, and 7/8; JNK1, 2, and 3; and p38-α,-β,-γ,and-δ [5,6]. diseases and several types of cancer [1,2]. The fundamental process Each MAPK has been identified for its action in independent signaling of apoptosis is conserved in different species, including yeasts, worms, pathways leading to pleiotropic functions [7].Theseproteinsare mice, and humans [3]. Many studies have shown that all of these involved in the regulation of various biological processes, including organisms share similar apoptosis-regulating proteins and that the mitosis, proliferation, differentiation, and apoptosis [8,9].Although apoptotic process proceeds through 2 main pathways, the cytoplasmic the mechanism of MAPK regulation of apoptosis varies with cell and and mitochondrial pathways. Both of these pathways are regulated by tissue type, generally, activation of JNK/SAPK and p38 MAPKs leads several signaling proteins, including the B-cell lymphoma 2 (Bcl-2) to induction of apoptosis, whereas p42/44 ERK activity is associated family proteins, kinases, and caspases [1]. with suppression of apoptosis [7]. The anti-apoptotic effect of ERK1/2 The mitogen-activated protein kinase (MAPK) pathway, an important can be regulated by a cascade comprised of the Ras/Raf/MEK (MAPK/ apoptosis-regulating pathway, has been shown to regulate apoptosis ERK kinase) 1/2/ERK1/2 pathways [8,9]. induced by UV irradiation, tumor necrosis factor (TNF), or TNF receptor Bax inhibitor-1 (BI-1), also known as TEGT (testis enhanced transcript), is an anti-apoptotic protein, which has been reported to suppress cell death induced by Bax expression or other diverse stress ⁎ Corresponding author at: Department of Animal Biotechnolgy, Konkuk University, Seoul 143-701, South Korea. Tel.: +82 2 450 4207; fax: +82 2 455 1044. stimuli [10,11]. Its anti-apoptotic function is widely conserved in most E-mail address: [email protected] (S.-G. Cho). eukaryotes, including yeasts and plants, and even in prokaryotes

0167-4889/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbamcr.2012.01.016 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888 877

[12–18], and recent studies have shown that BI-1 protects cells from Primers for C-terminal truncation of BI-1 included 5′-GGGAAGAATT- the reduction of calcium content in the (ER), CATGAACATATTTGATCGA-3′ and 5′-GGGAACTCGAGTCACTAGGACCGG ischemia–reperfusion injury, and ROS accumulation following ER TACT TA-3′. DNA fragments containing BI-1 or BI-1 ΔCweresubcloned stress [19–21]. Analysis of the sequence of BI-1 has led into a pEF mammalian expression vector containing an oligonucleotide to the suggestion that BI-1 is an intracellular membrane protein con- encoding an HA tag upstream of BI-1 and BI-1 ΔC. For RNA interference taining 6 or 7 transmembrane domains, and that these types of in- of BI-1 expression, a double-stranded oligonucleotide was designed tracellular membrane protein may have very important signaling and according to a common BI-1 sequence (Target sequence: 5′-CCCCGU- transport functions [22,23]. As a large number of disease-related muta- CAACGCAGCAGCA-3′ as previously described [26]) to allow formation tions have been found in intracellular membrane proteins and are of the hairpin structure in the expressed oligo-RNA. This was then reported to affect their stability, folding, and recognition in the mem- cloned into the pSilencer vector from Ambion (Austin, TX). ERK-1 and brane bilayer, intracellular membrane proteins, including BI-1, may be 2specific siRNAs (Cat. No. sc-29308) and control siRNA purchased important targets for therapeutic applications [24,25]. from Santa Cruz were used for RNA interference of ERK expression, Here, we studied the anti-apoptotic effects of several BI-1 family according to the manufacturer's protocol. proteins containing BI-1-like transmembrane domains and attempted to reveal the molecular mechanism of BI-1-mediated anti-apoptosis using cells and transgenic (TG) mice overexpressing BI-1. 2.4. Reverse transcriptase PCR (RT-PCR)

2. Materials and methods Total RNA was isolated from different tissues using TRIzol from Invitrogen (Carlsbad, CA) according to the manufacturer's instructions. 2.1. Antibodies and materials The first-strand cDNA was prepared from 2ug of total RNA isolated using reverse transcriptase with oligo-dT primer. cDNA was subjected Chemicals including etoposide, 2′,7′-dichlorodihydrofluorescein to PCR amplification with DNA primers selective for . PCR diacetate (H2-DCFDA), and N-acetylcysteine were purchased from was carried out using Hot-start Taq polymerase (Finnzymes, Espo, BioMOL (Plymouth Meeting, PA) and electrophoresis reagents Finland), as follows; for an initial denaturation step at 95 °C for and protein assay kits were from Bio-Rad (Hercules, CA). The primary 15 min, followed by 35 amplification cycles at 95 °C for 30 s, 51–59 °C antibodies used included those against hemagglutinin (HA, Cat. for 30 s and 72 °C for 30 s, and a final step at 72 °C for 5 min. All samples No. SC-805), green fluorescence protein (GFP, Cat. No. SC-8334), were run in triplicate. Primer sequences were as follows: TMBIM4 [F]: actin (Cat. No. SC-8432), cytochrome C (Cat. No. SC-13156), 5′-CATCCGAATGGCCTTTCTG-3′,TMBIM4[R]:5′-GTAGGTACAGGTT- voltage-dependent anion channel (VDAC, Cat. No. SC-8828), ERK AAGGGG-3′; FAIM2 [F]: 5′-GTACTGGGCATCCTATGCT-3′, FAIM2 [R]: (Cat. No. SC-153), JNK (Cat No. SC-7345), p38 (Cat. No. SC-7149), 5′-GTACTGGGCATCCTATGCT-3′; GHITM [F]: 5′-CAAGAATTGGGATCC- phospho-JNK (Cat. No. SC-6254), and poly (ADP-ribose) polymerase GGCG-3′, GHITM [R]: 5′-CCTTGACATACTGAGGCC-3′;BI-1[F]:5′- (PARP, Cat. No. SC-7148) (Santa Cruz Biotechnology, Santa Cruz, CA), GCAGGGGCCTATGTCCAT-3′,BI-1[R]:5′-AGGATGCTGGGGTTGACAGC- and phospho-p38 (Cat. No. 9211S), and phospho-ERK (Cat. No. 9106S) 3′; GFP-exo [F]: 5′-CTGAGCAAAGACCCCAACG-3′. The integrity of each (Cell Signaling, Danvers, MA). RNA sample was checked by RT-PCR with mouse glyceraldehyde- 3-phosphate dehydrogenase (GAPDH) primers (GAPDH [F]: 5′-AGAA- 2.2. Analysis of the amino acid sequences of BI-1 family proteins CATCATCCCTGCATCC-3′, GAPDH [R]: 5′-CACCACCTTCTTGATGTC-3′)as an internal standard. Amino acid sequences of BI-1 family proteins were retrieved from GenBank (NCBI). Accession numbers of the protein sequences are as follows: TMBIM4 (transmembrane Bax inhibitor motif containing 4: 2.5. Cell culture and DNA transfection NP_057140.1), FAIM2 (Fas apoptotic inhibitory molecule 2: NP_036438.2), GHITM (growth hormone-inducible transmembrane Human embryonic (HEK) 293 cells were cultured at 37 °C in protein: NP_055209.2), and BI-1 (Bax inhibitor-1: NP_003208.1). Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% The degree of sequence identity was determined using expected fetal bovine serum (FBS; Hyclone, Logan, UT) and 100 U/ml penicillin/ value calculation based on the GenBank division. Multiple sequence 100 μg/ml streptomycin (P/S). For DNA transfection, HEK 293 cells alignment was performed using the ClustalW program, and the were incubated overnight at a density of 3×106 cells per 100-mm SMART domain analysis program (http://smart.embl-heidelberg.de/) culture dish and transfected with the expression vectors indicated was used for prediction of the domain architecture of BI-1 family pro- using Lipofectamine (Invitrogen). For selection of HEK 293 cells that teins. ScanProsite (http:www.expasy.org/tools/scanprosite/PROSITE) stably expressed BI-1 or BI-1 ΔC, cells were transfected with 6 μg and NCBI RPS-BLAST (http://www.ncbi.nlm.nih.gov/Structure/cdd/ pEF HA (control), pEF HA-BI-1, or BI-1 ΔC and treated with G418 wrpsb.cgi) were used for conserved domain database screening. (200 μg/ml) 24 h after transfection. After selection with G418, the surviving HEK 293 cells were analyzed by RT-PCR and western 2.3. Plasmid construction blotting.

BI-1 family genes were amplified by PCR using human fetal brain cDNA (BD Biosciences, Franklin Lakes, NJ) and the following 2.6. Determination of cell viability primers: BI-1 (F) (5′-GGGAAGAATTCATGAACATATTTGATCGA-3′), BI-1 (R) (5′-GGGAACTCGAGTCATTTCTTCTCTTTCTT-3′), GHITM (F) Cells were plated at a density of 5×104 cells in 96-well plates and (5′-ATGTTGGCTGCAAGG CTG-3′), GHITM (R) (5′-TGATCCCATTAACTC- cell viability was evaluated via a conventional 3-(4,5-Dimethylthiazol- GATGCTGA-3′), FAIM2 (F) (5′-ATGACCCAGGGAA A-3′), FAIM2 (R) 2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction assay. Briefly, (5′-CACTAACCGAGAATGA-3′), TMBIM4 (F) (5′-ATGTGGTCGAGTGTA-3′), after incubation, the cells were treated with MTT solution (final concen- and TMBIM4 (R) (5′-GTTAATAAAAAGTAA-3′). The PCR products tration, 0.25 mg/ml) for 4 h at 37 °C. Dark blue formazan crystals that were digested with EcoRI/XhoI, subcloned into pDS-GFP, and the formed in the intact cells were dissolved with DMSO and absorbance plasmid sequence was then confirmed by DNA sequencing. Using was then measured at 570 nm using an ELISA reader (BD Biosciences). PCR-based methods, a mutant form (BI-1 ΔC) of BI-1 was generated Results were expressed as a percentage of MTT reduction, and absor- by deletion of the last 9 amino acids at the protein's C-terminal region. bance exhibited by the control cells was arbitrarily set as 100%. 878 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888

2.7. DAPI staining incubated for 1 h with fluorescein isothiocyanate (FITC)-conjugated secondary antibody (1:200). Immunofluorescent images were ana- Cells were collected by centrifugation, washed once with cold PBS, lyzed using confocal microscopy (Nikon C1 laser scanning confocal fixed in ice-cold methanol/acetic acid (1:1, v/v) for 5 min, and then microscope). stained with 0.8 mg/ml of 4′,6-diamidino-2-phenolindole (DAPI) in the dark. Morphological changes in apoptotic cells were assessed vi- sually under a Zeiss Axiovert 200 microscope from Carl Zeiss (Jena, 2.13. Generation of BI-1 TG mice Germany) in the region of DAPI fluorescence (excitation, 351 nm; emission, 380 nm). The standard microinjection procedure was used for the production of TG mice (Macrogen, Seoul, Korea). Briefly, fertilized mouse eggs were 2.8. Determination of mitochondrial membrane potential flushed from the oviducts of superovulated C57BL/6NCrjBgi mice and male pronuclei were injected with pEF HA/BI-1 DNA (2 ng/μl) that Cells were collected after exposure to 30 μM etoposide for 24 h. had been linearized with PI–SceI at its cognate restriction site located They were washed and resuspended in 1 ml PBS containing 2 μM in the pEF HA vector. The injected eggs were reimplanted in the

DiOC6 and incubated at 37 °C in the dark for 15 min. Mitochondrial oviducts of pseudo-pregnant imprint control region (ICR) recipient membrane potential (Δψm) was assessed using a FACSCalibur flow females. At 3 weeks of age, the animals were tested for the presence cytometer from BD Biosciences. of the transgene by PCR analysis of their genomic DNA. For genomic DNA-PCR analysis, primer pairs containing the BI-1 coding sequence 2.9. Subcellular fractionation (BI-1 [R]: 5′-CATGAAAGCAGTGGGAAGGATG-3′) were used together with the elongation factor (EF) promoter sequence (promoter [F]: Cells were harvested in an isotonic mitochondrial buffer (210 mM 5′-GATGGAGTTTCCCCACACTGA-3′). mannitol, 70 mM sucrose, 1 mM EDTA, and 10 mM HEPES, pH 7.5) supplemented with the protease inhibitor cocktail Complete (Roche, Mannheim, Germany) and homogenized for 30–40 strokes using a 2.14. Isolation and culture of mouse embryonic fibroblast (MEF) or Dounce homogenizer. To eliminate nuclei and unbroken cells, samples splenocyte cells were centrifuged at 500 ×g for 5 min at 4 °C. The supernatant was removed and centrifuged at 10,000×g for 30 min at 4 °C to obtain a In order to isolate MEFs, 14 day-7 embryos from control and BI-1 heavy membrane pellet enriched with mitochondria and the resulting TG mice were used as previously described [27]. Briefly, placental supernatant was stored as the cytosolic fraction. membranes, amniotic sac, head, and primordial blood organs were removed. The remaining carcass was washed and minced in 2 ml 2.10. Western blot analysis PBS using a syringe and an 18-gauge needle, passed through a 100-μm strainer for removal of large fragments, and placed in a 60-mm dish Cells in 60-mm dishes were washed 3 times in ice-cold PBS, containing DMEM, 10% FBS, 100 U/ml penicillin, and 100 μg/ml strepto- scraped from the dishes, and collected in extraction buffer (1% Triton mycin. MEF culture used for each experiment was derived from a single X-100, 100 mM Tris–HCl, p 7.5, 10 mM NaCl, 10% glycerol, 50 mM sodium embryo. Cells were passaged at 106 per 100 mm dish when they fluoride, and 1 mM phenylmethylsulfonyl fluoride [PMSF]). After the cells reached ~80% confluence. Spleens were surgically removed and kept had been incubated on ice for 30 min, the lysates were centrifuged and on ice in sterile DMEM supplemented with 1% FBS. Bulk cultures were the amount of protein in the cleared lysates was quantified using the established in DMEM supplemented with 10% FBS, 2 mM L-glutamine, Bradford Protein Assay Reagent (Bio-Rad). Equal amounts of protein and 100 U/ml P/S. were separated on 10%–12% SDS PAGE gels and transferred electropho- retically onto nitrocellulose membranes (0.2 mm). These membranes were blocked with 3%–5% non-fat dry milk and Tris-buffered saline 2.15. Histological analysis (TBS) and subsequently probed with primary antibody in TBS containing 3% non-fat dry milk. Antibody–antigen complexes were detected using Dissected spleen from control or BI-1 TG mice was fixed in 4% goat anti-mouse IgG or goat anti-rabbit IgG peroxidase conjugates, paraformaldehyde overnight at 4 °C, washed, and stored in 70% ethanol followed by use of an enhanced chemiluminescence (ECL) detection until ready for processing for paraffin embedding. Embedded tissue was kit from Amersham (Buckinghamshire, UK). sectioned (thickness, 5 μm) on a Jung RM 2025 microtome from Leica. For histological examination, the slides were stained with hematoxylin 2.11. Detection of intracellular ROS production and eosin (H&E) according to the standard protocols. For immunohisto- chemistry, sections were then incubated with the respective antibodies For measurement of intracellular ROS levels, cells were incubated in 0.5% blocking buffer (normal goat serum). Sections were washed and with 10 μM H2-DCFDA for 30 min at 37 °C in the dark, washed, and incubated with anti-mouse secondary antibodies (Oncogene Research suspended in PBS. Fluorescence-stained cells were imaged on an Products, San Diego, CA) diluted 1:1000 in 0.5% blocking buffer for inverted fluorescence microscope and analyzed with a FACSCalibur 1 h at room temperature. Vectastain Elite ABC and diaminobenzidine flow cytometer using CellQuest 3.2 software for acquisition and (DAB) substrate kits were used to detect immuno-peroxidase staining analysis. according to the manufacturer's instructions (Vector Laboratories, Burlingame, CA). 2.12. Immunofluorescence analysis

The cells were treated with MitoTracker red dye (Invitrogen, 2.16. Statistical analysis 25 nM) for 15 min in an incubator (37 °C, 5% CO2), washed three times with PBS, fixed for 15 min with 4% cold paraformaldehyde All values are expressed as the mean±SE. Each value is the mean of and rinsed three times with PBS. Cells were incubated for 1 h with at least 3 independent experiments in each group. Statistical significance 5% normal goat serum (Vector Laboratories, Burlingame, CA) to of the differences between the 2 cell populations was determined using block nonspecific staining, followed by incubation with anti-HA the two-tailed Student's t-test (Origin) and P values equal to or less than antibodies overnight at 4 °C. After washing with PBS, cells were 0.05 were considered significant. J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888 879

Fig. 1. Bioinformatic analysis revealed BI-1 family proteins. (A) Amino acid alignment of BI-1 family proteins was conducted. Protein sequences of TMBIM4, FAIM2, GHITM, and BI-1 were aligned using ClustalW. Identical amino acids highlighted in black and gray indicate conservative changes. (B) Phylogenetic tree analysis was carried out for selected members of the BI-1 family of proteins. The tree was derived from neighbor-joining distance analysis using MEGA4. (C) Protein domain and topology prediction was conducted for BI-1 family proteins. UPF0005 (uncharacterized protein family 0005) domain and transmembrane architecture of BI-1 family proteins were derived from TMHMM, TMPRED, and SMART domain analysis. 880 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888

Fig. 2. BI-1 displayed effective protection against etoposide-induced apoptosis. (A) An expression assay was performed for BI-1 family proteins. HEK 293 cells were transiently transfected with pDS-GFP-GHITM, pDS-GFP-TMBIM4, pDS-GFP-FAIM2, pDS-GFP-BI-1, or control vector using Lipofectamine. Using specific primers (BI: BI-1, FA: FAIM2, TM4: TMBIM4, GH: GHITM), RT-PCR analysis demonstrated endogenous or exogenous BI-1 family mRNA levels in each transfectant. Expression of the specific proteins was checked via western blot analysis with GFP antibody. Actin was used as a loading control. (B, C) Cell viability and apoptotic cell rates were measured using the MTT assay and DAPI staining analysis. Following transfection for 24 h with each vector, cells were treated with or without etoposide (7.5, 15, and 30 μM) for a further 48 h. All data are expressed as the means± SE of values obtained in 5 separate experiments. (D) Following treatment of transfectant cells with or without etoposide (30 μM) for 0, 24, or 48 h, whole cell extracts were analyzed for PARP cleavage with a specific polyclonal antibody.

3. Results from a base tree calculated by the neighbor-joining method using ClustalW (Fig. 1B). In the phylogenetic tree, these family proteins 3.1. Overexpression of BI-1 family proteins shows an anti-apoptotic effect were closely related and the relationship between family proteins on stress-induced cell death was paralogous. All of the proteins were analyzed to determine whether or not they contained conserved hydrophobic transmem- From the NCBI sequence database we retrieved amino acid se- brane domains known as UPF0005 (uncharacterized protein family quences for members of the BI-1 family proteins, such as TMBIM4, 0005); however, they were expected to have slightly different trans- FAIM2, and GHITM. Comparison of the primary structure of BI-1 membrane domain structures (Fig. 1C). We found that GFP-tagged and other BI-1 family proteins revealed 32% identity with TMBIM4, BI-1 family genes, such as FAIM2, TMBIM4, and GHITM had similar 31% identity with GHITM, and 25% identity with FAIM2, but no sig- expression levels in the respective transfected HEK 293 cells and nificant identity with GRINA (glutamate receptor, ionotropic, N- tested the ability of BI-1 family proteins to suppress stress- methyl D-aspartate-associated protein 1) and TMBI1 (transmem- induced apoptosis (Fig. 2A). Although all of the BI-1 family pro- brane Bax inhibitor motif containing 1) (Fig. 1). Homologous BI-1 teins showed a suppressive effect on etoposide-induced cell family proteins were aligned using a multiple sequence alignment death, overexpression of BI-1 showed the most significant anti- tool (ClustalW) and we determined that the hydrophobic and apoptoticreactioninHEK293cells(Fig. 2BandC).Next,we polar residues were preserved in specific positions (Fig. 1A). Based checked PARP cleavage after etoposide treatment and found that on the sequences of BI-1 family proteins, a phylogenetic tree was overexpression of BI-1 could significantly block stress-induced PARP constructed by the maximum likelihood method using MEGA4 cleavage (Fig. 2D). J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888 881

Fig. 3. BI-1 overexpression led to prevention of mitochondria-mediated-apoptosis and ROS production in etoposide treated cells; however, BI-1 ΔC overexpression did not. (A) The construct of the deletion mutant of BI-1 (BI-1 ΔC) is shown in the left panel. Expression of HA/BI-1 or HA/BI-1 ΔC was confirmed via RT-PCR using HA/BI-1 primers and western blot analysis with HA antibody as a control, BI-1, or BI-1 ΔC cells (right panel). Actin was used as a loading control. (B) Cell viability was measured with the MTT assay. Stable transfectant cells were treated with etoposide (7.5, 15, or 30 μM) for 48 h. (C) Apoptotic cells were assessed by DAPI staining. Control, BI-1, or BI-1 ΔC cells were treated with or without etoposide (7.5, 15, or 30 μM) for 48 h. Cells with apoptotic nuclear morphology were counted via fluorescence microscopy after DAPI staining. Results represent the means±SE of values obtained from 3 separate experiments. (D) Mitochondrial membrane depolarization was quantified by DiOC6 staining and subsequent FACS analysis. The cells were incubated in 30 μM etoposide for 24 h, then stained with 2 μMDiOC6, and analyzed by FACScan. (E) Following treatment with 30 μM etoposide for 24 h, samples were harvested in isotonic mitochondrial buffer and homog- enized. As described in the Materials and methods section, supernatant and mitochondrial pellets corresponding to 30 μg mitochondrial proteins were analyzed by western blotting.

(F) Following treatment with 30 μM etoposide for 36 h, cells were stained with the fluorescent probe H2-DCFDA (10 μM), and intracellular ROS levels were measured by fluorescence microscopy or flow cytometry. Increased ROS level was indicated by an increase in fluorescence. All data are expressed as the means±SE of values obtained in 3 separate experiments. (G) Immunofluorescence of HEK 293 cells transfected with plasmids containing GFP- or HA-tagged BI-1 and BI-1 ΔC. The green channel shows the fluorescence of GFP or HA fusion proteins and the red channel shows the fluorescence of the MitoTracker indicating the subcellular patterns of mitochondria. 882 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888

3.2. BI-1 overexpression leads to inhibition of mitochondria-mediated the C-terminal region, downstream of the UPF domain of BI-1 and apoptosis tested the effect of expression of BI-1 or BI-1 ΔC on apoptotic cell death (Fig. 3A). Compared with the control cells, BI-1-expressing We prepared BI-1 ΔC, which was known to be an inactive mutant cells showed significantly greater resistance to etoposide-induced form of BI-1 [11,28,29], by deleting 9 amino acids (EKDKKKEKK) at cell death, along with a significant increase in cell viability, but BI-1 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888 883

ΔC-expressing cells did not (Fig. 3B). Etoposide-induced nuclear confirmed by the use of ERK-specific siRNA. Suppression of ERK acti- fragmentation was apparently reduced by BI-1 overexpression, but vation by knockdown of ERK expression by ERK-specific siRNA led to not by BI-1 ΔC overexpression (Fig. 3C). Next, we examined the effects the apparent downregulation of ERK phosphorylation and significant of BI-1 overexpression on both loss of mitochondrial membrane abolishment of the anti-apoptotic function of BI-1 expression (Fig. 4C potential (Δψm) and release of cytochrome C. Triggering apoptosis and D), strongly suggesting that BI-1-induced sustained ERK activa- with etoposide led to perturbation of the mitochondrial membrane tion may be involved in the anti-apoptotic effect of BI-1. potential in control and BI-1 ΔC cells, whereas BI-1 overexpression apparently suppressed the decrease in mitochondrial potential 3.4. BI-1-mediated ERK activation leads to downregulation of (Fig. 3D). The release of mitochondrial cytochrome C following ROS-associated apoptosis etoposide stimulation was also inhibited by BI-1 overexpression, but not by BI-1 ΔCexpression(Fig. 3E). The purity of the cytosolic As intracellular ROS production was analyzed by incubation of and mitochondrial fractions was confirmed by immunoblotting for the cells with H2-DCFDA, we were able to determine that in- VDAC. Next, for assessment of overall ROS levels in BI-1-expressing creased ROS production following etoposide treatment was signif- cells, the cell-permeable ROS indicator, carboxy-H2-DCFDA, was icantly suppressed by BI-1 overexpression, but not by BI-1 ΔC used after treatment of the cells with etoposide. Etoposide treatment expression (Fig. 5A). Pretreatment of cells with the cell-permeant strongly induced the generation of fluorescent 2′,7′-dichlorofluorescein ROS scavenger, N-acetylcysteine (NAC), led to downregulation of ROS (DCF) in control cells. However, BI-1-expressing cells showed low-level production and, interestingly, pretreatment with the pharmacological generation of fluorescent DCF (Fig. 3F), implying that BI-1 overexpres- MEK inhibitor, PD98059, resulted in upregulation of ROS generation sion resulted in greater resistance to mitochondria-mediated apoptosis and abolishment of the ROS-suppression effect of BI-1 overexpression, and ROS generation. To determine the effect of the C-terminal deletion suggesting an adverse effect of ERK activity on etoposide-induced of BI-1 in BI-1 cellular localization, plasmids containing GFP- or ROS production in HEK 293 cells. Next, we analyzed the effects of HA-tagged BI-1 and BI-1 ΔCwerepreparedandexpressedincells.As NAC pretreatment on ERK activation and PARP cleavage and found shown in Fig. 3G, the fluorescent images of BI-1 and BI-1 ΔC revealed that NAC pretreatment led to the apparent suppression of PARP a similar localization pattern. Our data are similar to results reported cleavage, but did not significantly influence ERK activation (Fig. 5B). in the previous study [13], indicating that C-terminal deletion did not Taken together, our results suggest that BI-1-induced ERK activation affect the cellular localization of BI-1. plays a key role in the regulation of intracellular ROS production and that suppression of ROS production by the BI-1/ERK activation 3.3. ERK pathway plays an important role in the anti-apoptotic function pathway may be a significant factor in the anti-apoptotic effect of of BI-1 BI-1.

As the MAPKs are a family of proteins that mediate distinct signal- 3.5. BI-1 overexpression in BI-1 TG mice results in anti-apoptotic effects, ing cascades in the regulation of apoptosis, we tested the effect of splenomegaly, and extramedullary megakaryopoiesis MAPK activity on the anti-apoptotic function of BI-1. First, BI-1- or BI-1 ΔC-overexpressing cells were treated with etoposide and apo- Our data suggest that the ERK pathway executes an important ptosis induction was monitored at several time points by examination function in the anti-apoptotic and anti-oxidative effects of BI-1. of PARP cleavage. We were able to detect apparent PARP cleavage in To confirm the physiological function of the BI-1/ERK pathway in response to etoposide treatment in HEK 293/pEF HA control cells vivo, we generated BI-1 TG mice by exogenously expressing HA- and significant suppression of PARP cleavage in BI-1-overexpressing tagged BI-1 on the EF promoter. The BI-1 transgene was transferred cells (Fig. 4A). As MAPK activation was monitored at several time into mice by pronuclear injection and transgene integration was points after etoposide treatment, we found that control and BI-1 detected in the genomic DNA of several F0 pups (Fig. 6A). We ΔC-overexpressing cells showed a biphasic response to ERK activa- were also able to confirm that HA-tagged BI-1 protein was highly tion in a time-dependent manner. ERK phosphorylation peaked 24 h expressed in diverse TG mouse tissues, including , lung, after etoposide treatment and returned to basal levels after 48 h. On , spleen, and testis (Fig. 6B). Next, we prepared MEF cells the other hand, in BI-1-overexpressing cells, ERK was strongly phos- from BI-1 TG mice and analyzed the effect of BI-1 overexpression. phorylated even after 24 h of etoposide treatment and its phosphory- First, through RT-PCR and western blot analysis, we confirmed ex- lation increased thereafter over a period of 48 h. However, BI-1 pression of the BI-1 transgene in BI-1 TG MEF cells (Fig. 6C). Fol- overexpression had little or no influence on the activation of either lowing etoposide treatment, MEF cells from BI-1 TG mice showed JNK or p38 MAPK. In order to further confirm the direct role of BI-1 significantly lowered PARP cleavage, consistent with the data expression on etoposide-induced ERK activation, BI-1-specific obtained with BI-1-expressing HEK 293 cells (Fig. 6D). Compared shRNA was introduced into cells overexpressing BI-1 or BI-1 ΔC. with the control MEF cells, we were also able to detect higher Knockdown of BI-1 expression led to a dramatic increase in ERK activation and decreased ROS levels in BI-1 TG MEF cells, and etoposide-induced PARP cleavage in controls, as well as cells overex- knockdown of BI-1 expression by BI-1-specific shRNA led to abol- pressing BI-1 and BI-1 ΔC(Fig. 4B). Importantly, knockdown of BI-1 ishment of BI-1-induced ERK activation and downregulation of in- expression resulted in the abolishment of upregulation of ERK activa- tracellular ROS production in MEF cells (Fig. 6D and E). Moreover, tion in BI-1-overexpressing cells, supporting the notion that BI-1 spe- the results obtained in PD98059- or NAC-pretreated BI-1 TG MEF cifically leads to sustained ERK activation. Involvement of sustained cells supported our hypothesis that BI-1-induced ERK activation ERK activation in the anti-apoptotic function of BI-1 was also may play an important role in the regulation of intracellular ROS

Fig. 4. Anti-apoptotic effect of BI-1 was closely associated with activation of ERK1/2. (A) Time-dependent induction of apoptosis and ERK1/2 phosphorylation was analyzed based on etoposide stress. Control, BI-1, or BI-1 ΔC cells were treated with 30 μM etoposide for 24, 36, or 48 h, and protein extracts were prepared. Cellular proteins were analyzed by western blotting with anti-PARP, anti-phospho-ERK (p-ERK), anti-total ERK (ERK), anti-phospho-JNK (p-JNK), anti-total JNK (JNK), anti-phospho-p38 (p-p38), or anti-total p38 (p38) specific antibody. Actin was used as a loading control. Representative blots from 3 independent experiments are presented. (B) Stable cells were transiently transfected with pSilencer (control) and pSilencer/shBI-1 plasmids. The expression level of BI-1 was detected by western blotting in the cells indicated, with or without treatment with etopo- side (30 μM and 36 h). Equal protein loading was shown by reprobing of the blot with the actin antibody. Phosphorylation of ERK1/2 and total ERK1/2 was examined by western analysis in the same cells. (C and D) Transfectant cells were transfected with scramble (−) or ERK-specific siRNA (+), and were further cultured with 30 μM etoposide for 36 h. Cells were lysed and subjected to western blotting with anti-PARP or anti-p-ERK antibody. The membrane was sequentially reprobed with anti-ERK or anti-actin, as indicated (C), and apoptotic cells were assessed by FACS analysis using PI staining (D). 884 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888

Fig. 5. BI-1 overexpression led to suppression of etoposide-induced ROS generation, and was dependent on ERK activation. Cells were cultured in the presence or absence of etoposide (30 μM), NAC (1 mM), and PD98059 (20 μM) for 24 h. NAC and PD98059 were added at the beginning of the experiment. (A) The level of intracellular ROS in stable transfectant cells was visualized by DCF staining. ROS levels were visualized by fluorescence of the dye H2-DCFDA (10 μM) and quantified with FACScan. The results shown are representative of 3 independent experiments performed in triplicate. (B) The cell lysates were subjected to western blotting with antibody against p-ERK, total ERK, actin, or PARP. generation and etoposide-induced apoptosis. We also analyzed the family proteins and have cytoprotective effects, but their precise in vivo effect of BI-1 overexpression in BI-1 TG mice and observed role and their regulatory mechanisms remain to be determined that the spleen of BI-1 TG mice was enlarged, with an average [30–32]. In order to gain insight into the significance of BI-1 family weight of approximately 0.2 g, which were 3–4 times heavier proteins in apoptosis, we first investigated the structure of several than that of control mice (Fig. 7A). Morphological and histochemi- BI-1 family proteins, including FAIM2, TMBIM4, and GHITM. Bioin- cal analysis of the enlarged spleens indicated that the splenomega- formatic analysis revealed that BI-1 family proteins share a func- ly was accompanied by disruption of the normal architecture of tionally unknown transmembrane protein domain (UPF0005 spleen pulp, resulting in loss of the boundaries (marginal zones) motif), with variable N-terminal and C-terminal regions and 6 or that separate white pulp from red pulp (Fig. 7B). Interestingly, we 7 transmembrane passages (Fig. 1). These morphological resem- detected a significant increase in megakaryocytes in the spleen blances or variations between BI-1 family proteins may determine from BI-1 TG mice and, in fact, we could observe a consistent in- their cellular functions. Previous studies have already shown that crease in the level of ERK phosphorylation in the BI-1 TG spleen tis- FAIM2 could protect cells against Fas-induced apoptosis, and sue (Fig. 7C). Moreover, from the analysis of splenocytes prepared TMBIM4 was also shown to inhibit both intrinsic and extrinsic from BI-1 TG mice, we could confirm that BI-1 overexpression pathway-mediated apoptosis [33,34]. Moreover, GHITM has been plays an important role in etoposide-induced apoptosis of spleno- reported to play a role in the maintenance of mitochondrial mor- cytes, which is apparently modulated by a BI-1-mediated ERK path- phology and delay of apoptotic release of cytochrome C [35]. way (Fig. 7D and E), suggesting that BI-1-induced ERK activation Studies have also been carried out to determine whether BI-1 may be a significant signaling module in the anti-apoptotic and im- plays a cytoprotective role against ER stress, which was induced mune response-regulating functions of BI-1. by pharmacological inducers such as tunicamycin and thapsigargin [19]. These results suggest that BI-1 family proteins may play a 4. Discussion centralroleinanti-apoptoticregulationinanimals.Deficiency of BI-1 in mice has resulted in elevated levels of ER stress markers The apoptotic process is regulated by a variety of diverse pro- and the activation of stress kinases, such as JNK and p38 MAPK teins, including Bcl-2 and caspase family proteins, and an under- [21]. In this study, we compared the anti-apoptotic properties of standing of their functions is important for the elucidation of the BI-1 family proteins against stress-associated or mitochondria- regulatory mechanisms of many apoptosis-related diseases [1,2]. mediated apoptosis and found that BI-1 has a strong anti- The BI-1 family proteins containing BI-1-like transmembrane do- apoptotic effect on etoposide-induced apoptosis in HEK 293 cells mains were expected to play an important role in apoptosis. (Figs. 2 and 3). Generally, the MAPK pathway is important in the Some of these proteins have been reported to interact with Bcl-2 regulation of various cellular functions, including proliferation, J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888 885

Fig. 6. MEF cells from BI-1 TG mice apparently revealed ERK activation-dependent anti-apoptosis. (A) BI-1 transgene integrations in genomic DNA (10 μg) of representative F0 pups (TG-1 to TG-4) and background strain C57BL/6N (Control) were analyzed via genomic DNA-PCR analysis. We used specific primer pairs containing the BI-1-specific coding sequence (reverse primer) along with the EF promoter sequence (forward primer). Tail biopsies were used for genomic DNA-PCR analysis (lower panel). M: DNA size marker. (B) BI-1 expression levels were shown in diverse tissues of 6-month-old controls (C) or BI-1 transgenic (BI) mice (He: Heart, Lu: Lung, Li: Liver, Sp: Spleen, Te: Testis, St: Stomach). (C) Using a specificBI-1primer,we tested expression of BI-1 mRNA in each MEF cell from BI-1 TG mice and control mice (left panel). BI-1 expression was analyzed by western blotting using HA antibody in control and BI-1 TG MEF cells (right panel) (P/C: Positive control, normal MEF cells transiently transfected with BI-1). (D and E) Each MEF cell line was treated with etoposide in the absence or presence of PD98059, NAC, or BI-1-specific shRNA for 36 h. Cells were lysed, and the total lysates were subjected to immunoblotting for detection of phosphorylated ERK (p-ERK), total ERK (ERK), PARP, or actin (D). After treatment of etoposide for 24 h, cells were stained with DCF and analyzed using FACS analysis (E). differentiation, apoptosis, development, growth, and inflammation. significant induction in the PARP cleavage and ROS production of In mammalian cells, the MAPK family has been divided into 3 normal MEF cells. We think that BI-1-knockdown MEFs displayed groups, making up the ERK, JNK, and p38 pathways. Previous re- normal sensitivity to etoposide-induced apoptosis. Our result was ports have shown that BI-1 regulates ER stress markers, including similar with other previous study [19] showing that BI-1-deficient JNK phosphorylation in cells overexpressing or deficient in BI-1 MEF showed no significant difference in apoptotic ratio compared [36,37].However,inthisstudy,althoughlevelsofJNKandp38 with wild-type MEF. The causes of these results certainly not yet phosphorylation were not apparently changed on BI-1 overexpres- well understood, so we need further study. sion, BI-1 overexpression led to significant inhibition of stress- Interestingly, TG mice overexpressing BI-1 showed higher ERK ac- induced apoptosis via activation of the ERK pathway. Moreover, tivation and spontaneously developed splenomegaly and megakaryo- we attempted to reveal the molecular mechanism of BI-1- poiesis, which is a similar phenotype to that of certain transgenic mediated anti-apoptosis using cells and transgenic (TG) mice over- mice in which the ERK pathway has been upregulated or apoptosis expressing BI-1. In BI-1 TG MEF cells, knockdown of BI-1 expres- blocked [38–43]. These observations indicate that BI-1 overexpres- sion by BI-1-specificshRNAledtoenhancedPARPcleavage, sion may modulate the apoptotic signaling pathway and autoimmu- abolishment of BI-1-induced ERK activation and downregulation nity through regulation of ERK activation. Consistently, previous of intracellular ROS production in MEF (Fig. 6C).However,BI-1 studies also reported the involvement of ERK1/2 activation in the re- shRNA-induced inhibition of BI-1 expression did not result in a sistance to apoptosis in response to diverse apoptotic stimuli [44,45]. 886 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888

Fig. 7. BI-1 overexpression leads to splenomegaly and extramedullary megakaryopoiesis. (A) BI-1 TG mice were larger than control mice and displayed splenomegaly. Pictures are representative of 3 mice analyzed for each group. The histogram shows mean wet weight of spleens freshly dissected from adult control and BI-1 TG mice. (B) The spleens from control and BI-1 TG mice were fixed, embedded in paraffin, sectioned, stained with H&E, and examined by light microscopy. Megakaryocytes are indicated by black arrowheads. (C) Control and BI-1 spleen tissue was lysed, and an equal amount of cell lysate was blotted for p-ERK, ERK, and actin. Paraffin-embedded spleen tissues from each mouse were probed with anti-p-ERK antibody. Findings replicated in independent preparations of six pairs of mice. (D) BI-1 expression levels were observed in control and BI-1 TG splenocytes. (E) Control and BI-1 TG splenocytes treated with etoposide in the absence or presence of BI-1 shRNA, or ERK siRNA. Apoptotic cells were detected by PI staining using FACS analysis. All data are expressed as the means±SE of values obtained in 4 separate experiments.

Our study shows that BI-1 expression leads to sustained ERK activa- pathways by activation of calcium/calmodulin-dependent protein ki- tion during etoposide treatment, consistent with previous findings nase (CaMK) [48], and ERK activation by CaMKII was also demon- demonstrating the importance of sustained ERK1/2 activation in cell strated through direct binding of the protein with MEK1 or Raf-1 survival [46,47]. [49,50]. On the basis of these reports, research into the exact mecha- The mechanism by which BI-1 regulates ERK activation remains nism for the involvement of BI-1 in ERK activation is underway. In ad- unknown. ERK1/2 regulates diverse cellular processes, including cell dition, several reports have suggested that BI-1 homolog proteins cycle progression, proliferation, and differentiation. Moreover, it is could protect yeast and plant cells from oxidative stress [13–15,28]. implicated in immune system development, , and tissue In relation to this finding, we found that ERK activation by BI-1 over- development. Complex upstream cascades, including numerous expression led to a reduction in intracellular ROS levels (Figs. 5 and GTP-binding proteins, kinases, and phosphatases, have been reported 6). ROS are known to be generated by several pathways and produc- to influence ERK activation. A previous study reported that the C ter- tion of ROS results in cell damage, aging, and apoptosis [51,52]. Mito- minus of plant BI-1 interacts with calmodulin, which contains a chondria are the major sites of ROS production and mitochondrial calcium-binding domain that regulates diverse intracellular signal ROS leads to oxidative stress with enhanced activity of the J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888 887 antioxidant defense system [53,54]. Anti-apoptotic Bcl-2 protein is [12] M. Kawai-Yamada, L. Jin, K. Yoshinaga, A. Hirata, H. Uchimiya, Mammalian Bax-induced plant cell death can be down-regulated by overexpression of one of the defense systems associated with defense against oxidants Arabidopsis Bax Inhibitor-1 (AtBI-1), Proc. Natl. Acad. Sci. U. S. A. 98 (2001) and a change in the cellular ROS level [55,56]. According to several 12295–12300. previous studies, ROS could be an important mediator of apoptosis [13] M. Kawai-Yamada, Y. Ohori, H. Uchimiya, Dissection of Arabidopsis Bax inhibitor-1 fi suppressing Bax-, hydrogen peroxide-, and salicylic acid-induced cell death, Plant and MAPK regulation [57]. Consistent with these ndings, our Cell 16 (2004) 21–32. data showed that ROS regulation may be dependent upon BI-1- [14] D. Baek, J. Nam, Y.D. Koo, D.H. Kim, J. Lee, J.C. Jeong, S.S. Kwak, W.S. Chung, C.O. Lim, dependent ERK activation. J.D. Bahk, J.C. Hong, S.Y. Lee, M. Kawai-Yamada, H. Uchimiya, D.J. Yun, Bax-induced Importantly, our results suggest that BI-1 may affect the autoim- cell death of Arabidopsis is meditated through reactive oxygen-dependent and -independent processes, Plant Mol. Biol. 56 (2004) 15–27. mune response, as indicated by the results in BI-1 TG mice, showing [15] H. Matsumura, S. Nirasawa, A. Kiba, N. Urasaki, H. Saitoh, M. Ito, M. Kawai-Yamada, that overexpression of BI-1 led to abnormal spleen morphology, H. Uchimiya, R. Terauchi, Overexpression of Bax inhibitor suppresses the fungal – splenomegaly, and megakaryocytes (Fig. 7A and B). Many scientists elicitor-induced cell death in rice (Oryza sativa L) cells, Plant J. 33 (2003) 425 434. [16] M. Kawai, L. Pan, J.C. Reed, H. Uchimiya, Evolutionally conserved plant homologue have suggested that diverse mechanisms are essential for autoim- of the Bax inhibitor-1 (BI-1) gene capable of suppressing Bax-induced cell death mune responses and that many determinants could induce these re- in yeast(1), FEBS Lett. 464 (1999) 143–147. sponses through abnormal regulation of the apoptosis-regulating [17] P. Sanchez, M. de Torres Zabala, M. Grant, AtBI-1, a plant homologue of Bax inhibitor-1, suppresses Bax-induced cell death in yeast and is rapidly upregulated mechanisms [58]. Apoptosis-regulating proteins have been reported during wounding and pathogen challenge, Plant J. 21 (2000) 393–399. to be involved in numerous signaling pathways that appear to be nec- [18] N. Bolduc, M. Ouellet, F. Pitre, L.F. Brisson, Molecular characterization of two plant essary for the maintenance of tolerance and thus for healthy immune BI-1 homologues which suppress Bax-induced apoptosis in human 293 cells, – fi Planta 216 (2003) 377 386. systems [59]. Inef cient apoptosis may lead to a failure to delete self- [19] H.J. Chae, H.R. Kim, C. Xu, B. Bailly-Maitre, M. Krajewska, S. Krajewski, S. Banares, reactive lymphocytes and expansion of immature and functionally J. Cui, M. Digicaylioglu, N. Ke, S. Kitada, E. Monosov, M. Thomas, C.L. Kress, J.R. defective lymphocytes. In the case of Bax deficiency or Bcl-2 overex- Babendure, R.Y. Tsien, S.A. Lipton, J.C. Reed, BI-1 regulates an apoptosis pathway linked to endoplasmic reticulum stress, Mol. Cell 15 (2004) 355–366. pression, an imbalance of pro- or anti-apoptotic proteins has been [20] B.C. Westphalen, J. Wessig, F. Leypoldt, S. Arnold, A. Methner, BI-1 protects reported to alter lymphocyte development [41] or lymphocytic leuke- cells from oxygen glucose deprivation by reducing the calcium content of the mia [42]. A crucial role in the autoimmune response has been sug- endoplasmic reticulum, Cell Death Differ. 12 (2005) 304–306. gested for the MAPK cascade [60–62] and recent reports have stated [21] B. Bailly-Maitre, C. Fondevila, F. Kaldas, N. Droin, F. Luciano, J.E. Ricci, R. Croxton, M. Krajewska, J.M. Zapata, J.W. Kupiec-Weglinski, D. Farmer, J.C. Reed, Cyto- that ERK could be involved in lymphoproliferative disease [39] and protective gene bi-1 is required for intrinsic protection from endoplasmic polycythemia vera-like disease [40]. Some papers have suggested reticulum stress and ischemia-reperfusion injury, Proc. Natl. Acad. Sci. U. – that the ERK signaling pathway plays an important role in diverse au- S. A. 103 (2006) 2809 2814. [22] W. Yang, S.H. Xia, Mechanisms of regulation and function of G-protein-coupled toimmune responses, including T-cell or B-cell activation and lym- receptor kinases, World J. Gastroenterol. 12 (2006) 7753–7757. phoproliferation [39,63–66]. Therefore, further studies are required [23] J. Massague, R.R. Gomis, The logic of TGFbeta signaling, FEBS Lett. 580 (2006) to determine whether the autoimmune response is associated with 2811–2820. [24] E. Li, K. Hristova, Role of receptor tyrosine kinase transmembrane domains in cell regulation of apoptosis and ERK activation by BI-1 overexpression. signaling and human pathologies, Biochemistry 45 (2006) 6241–6251. Consequently, our results demonstrate that BI-1 expression may in- [25] V.M. Olkkonen, E. Ikonen, Genetic defects of intracellular-membrane transport, N. duce sustained ERK activation and reduce ROS production, leading Engl. J. Med. 343 (2000) 1095–1104. [26] M. Grzmil, P. Thelen, B. Hemmerlein, S. Schweyer, S. Voigt, D. Mury, P. Burfeind, Bax to suppression of autoimmune response-related apoptosis. inhibitor-1 is overexpressed in prostate cancer and its specific down-regulation by RNA interference leads to cell death in human prostate carcinoma cells, Am. J. Pathol. 163 (2003) 543–552. Acknowledgements [27] S. Parrinello, E. Samper, A. Krtolica, J. Goldstein, S. Melov, J. Campisi, Oxygen sensitivity severely limits the replicative lifespan of murine fibroblasts, Nat. This work was supported by a grant from the Korea Health 21 R&D Cell Biol. 5 (2003) 741–747. Project, Ministry for Health, Welfare and Family Affairs, Republic of [28] H.J. Chae, N. Ke, H.R. Kim, S. Chen, A. Godzik, M. Dickman, J.C. Reed, Evolutionarily conserved cytoprotection provided by Bax Inhibitor-1 homologs from animals, Korea (A08-4065 and A10-1712) and by National Research Foundation plants, and yeast, Gene 323 (2003) 101–113. (NRF) grant funded by the Korea government (MEST) (No. 2010- [29] H.R. Kim, G.H. Lee, K.C. Ha, T. Ahn, J.Y. Moon, B.J. Lee, S.G. Cho, S. Kim, Y.R. Seo, Y.J. 0020348). We acknowledge a graduate fellowship provided by the Shin, S.W. Chae, J.C. Reed, H.J. Chae, Bax Inhibitor-1 Is a pH-dependent regulator of Ca2+ channel activity in the endoplasmic reticulum, J. Biol. Chem. 283 Ministry of Education and Human Resources Development through (2008) 15946–15955. the Brain Korea 21 project, Republic of Korea. [30] J. Zhou, T. Zhu, C. Hu, H. Li, G. Chen, G. Xu, S. Wang, J. Zhou, D. Ma, Comparative ge- nomics and function analysis on BI1 family, Comput. Biol. Chem. 32 (2008) 159–162. [31] K. Reimers, C.Y. Choi, E. Mau-Thek, P.M. Vogt, Sequence analysis shows that References Lifeguard belongs to a new evolutionarily conserved cytoprotective family, Int. J. Mol. Med. 18 (2006) 729–734. [1] I.M. Ghobrial, T.E. Witzig, A.A. Adjei, Targeting apoptosis pathways in cancer [32] K. Reimers, C.Y. Choi, V. Bucan, P.M. Vogt, The Bax Inhibitor-1 (BI-1) family in therapy, CA Cancer J. Clin. 55 (2005) 178–194. apoptosis and tumorigenesis, Curr. Mol. Med. 8 (2008) 148–156. [2] S. Willis, C.L. Day, M.G. Hinds, D.C. Huang, The Bcl-2-regulated apoptotic pathway, [33] N.V. Somia, M.J. Schmitt, D.E. Vetter, D. Van Antwerp, S.F. Heinemann, I.M. Verma, J. Cell Sci. 116 (2003) 4053–4056. LFG: an anti-apoptotic gene that provides protection from Fas-mediated cell [3] M. van Gurp, N. Festjens, G. van Loo, X. Saelens, P. Vandenabeele, Mitochondrial death, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 12667–12672. intermembrane proteins in cell death, Biochem. Biophys. Res. Commun. 304 [34] C. Gubser, D. Bergamaschi, M. Hollinshead, X. Lu, F.J. van Kuppeveld, G.L. Smith, (2003) 487–497. A new inhibitor of apoptosis from vaccinia virus and eukaryotes, PLoS Pathog. [4] W. Zhang, H.T. Liu, MAPK signal pathways in the regulation of cell proliferation in 3 (2007) e17. mammalian cells, Cell Res. 12 (2002) 9–18. [35] T. Oka, T. Sayano, S. Tamai, S. Yokota, H. Kato, G. Fujii, K. Mihara, Identification of a [5] A.S. Dhillon, S. Hagan, O. Rath, W. Kolch, MAP kinase signalling pathways in novel protein MICS1 that is involved in maintenance of mitochondrial morphology cancer, Oncogene 26 (2007) 3279–3290. and apoptotic release of cytochrome C, Mol. Biol. Cell 19 (2008) 2597–2608. [6] E.K. Kim, E.J. Choi, Pathological roles of MAPK signaling pathways in human [36] G.H. Lee, H.K. Kim, S.W. Chae, D.S. Kim, K.C. Ha, M. Cuddy, C. Kress, J.C. Reed, H.R. diseases, Biochim. Biophys. Acta 1802 (2010) 396–405. Kim, H.J. Chae, Bax inhibitor-1 regulates endoplasmic reticulum stress-associated [7] T.G. Cross, D. Scheel-Toellner, N.V. Henriquez, E. Deacon, M. Salmon, J.M. Lord, reactive oxygen species and heme oxygenase-1 expression, J. Biol. Chem. 282 Serine/threonine protein kinases and apoptosis, Exp. Cell Res. 256 (2000) 34–41. (2007) 21618–21628. [8] G.L. Johnson, R. Lapadat, Mitogen-activated protein kinase pathways mediated by [37] H.R. Kim, G.H. Lee, E.Y. Cho, S.W. Chae, T. Ahn, H.J. Chae, Bax inhibitor 1 regulates ERK, JNK, and p38 protein kinases, Science 298 (2002) 1911–1912. ER-stress-induced ROS accumulation through the regulation of cytochrome P450 [9] C.J. Caunt, A.R. Finch, K.R. Sedgley, C.A. McArdle, Seven-transmembrane receptor 2E1, J. Cell Sci. 122 (2009) 1126–1133. signalling and ERK compartmentalization, Trends Endocrinol. Metab. 17 (2006) [38] T. Kamata, J. Hussain, S. Giblett, R. Hayward, R. Marais, C. Pritchard, BRAF 276–283. inactivation drives aneuploidy by deregulating CRAF, Cancer Res. 70 (2010) [10] R. Huckelhoven, BAX Inhibitor-1, an ancient cell death suppressor in animals and 8475–8486. plants with prokaryotic relatives, Apoptosis 9 (2004) 299–307. [39] M. Miyaji, R.L. Kortum, R. Surana, W. Li, K.D. Woolard, R.M. Simpson, L.E. Samelson, [11] Q. Xu, J.C. Reed, Bax inhibitor-1, a mammalian apoptosis suppressor identified by C.L. Sommers, Genetic evidence for the role of Erk activation in a lymphoproliferative functional screening in yeast, Mol. Cell 1 (1998) 337–346. disease of mice, Proc. Natl. Acad. Sci. U. S. A. 106 (2009) 14502–14507. 888 J.-H. Kim et al. / Biochimica et Biophysica Acta 1823 (2012) 876–888

[40] H. Akada, D. Yan, H. Zou, S. Fiering, R.E. Hutchison, M.G. Mohi, Conditional expression [53] S. Orrenius, V. Gogvadze, B. Zhivotovsky, Mitochondrial oxidative stress: implications of heterozygous or homozygous Jak2V617F from its endogenous promoter induces a for cell death, Annu. Rev. Pharmacol. Toxicol. 47 (2007) 143–183. polycythemia vera-like disease, Blood 115 (2010) 3589–3597. [54] C. Fleury, B. Mignotte, J.L. Vayssiere, Mitochondrial reactive oxygen species in cell [41] J.C. Rathmell, T. Lindsten, W.X. Zong, R.M. Cinalli, C.B. Thompson, Deficiency in death signaling, Biochimie 84 (2002) 131–141. Bak and Bax perturbs thymic selection and lymphoid homeostasis, Nat. Immunol. [55] K.A. Webster, R.M. Graham, J.W. Thompson, M.G. Spiga, D.P. Frazier, A. Wilson, N.H. 3 (2002) 932–939. Bishopric, Redox stress and the contributions of BH3-only proteins to infarction, [42] J.M. Zapata, M. Krajewska, H.C. Morse III, Y. Choi, J.C. Reed, TNF receptor-associated Antioxid. Redox Signal. 8 (2006) 1667–1676. factor (TRAF) domain and Bcl-2 cooperate to induce small B cell lymphoma/chronic [56] Z.X. Chen, S. Pervaiz, BCL-2: pro-or anti-oxidant? Front. Biosci. (Elite Ed.) 1 lymphocytic leukemia in transgenic mice, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) (2009) 263–268. 16600–16605. [57] A. Matsuzawa, H. Ichijo, Redox control of cell fate by MAP kinase: physiological [43] Y. Kozuma, S. Yuki, H. Ninomiya, T. Nagasawa, H. Kojima, Caspase activation is roles of ASK1-MAP kinase pathway in stress signaling, Biochim. Biophys. Acta involved in early megakaryocyte differentiation but not in platelet production 1780 (2008) 1325–1336. from megakaryocytes, Leukemia 23 (2009) 1080–1086. [58] J.B. Oliveira, S. Gupta, Disorders of apoptosis: mechanisms for autoimmunity in [44] Z. Lu, S. Xu, ERK1/2 MAP kinases in cell survival and apoptosis, IUBMB Life 58 primary immunodeficiency diseases, J. Clin. Immunol. 28 (Suppl. 1) (2008) (2006) 621–631. S20–S28. [45] J.W. Ramos, The regulation of extracellular signal-regulated kinase (ERK) in [59] E. Maniati, P. Potter, N.J. Rogers, B.J. Morley, Control of apoptosis in autoimmunity, mammalian cells, Int. J. Biochem. Cell Biol. 40 (2008) 2707–2719. J. Pathol. 214 (2008) 190–198. [46] Y.Y. He, J.L. Huang, C.F. Chignell, Delayed and sustained activation of extracellular [60] B. Kaminska, MAPK signalling pathways as molecular targets for anti- signal-regulated kinase in human keratinocytes by UVA: implications in carcino- inflammatory therapy—from molecular mechanisms to therapeutic benefits, genesis, J. Biol. Chem. 279 (2004) 53867–53874. Biochim. Biophys. Acta 1754 (2005) 253–262. [47] T. Nyunoya, M.M. Monick, L.S. Powers, T.O. Yarovinsky, G.W. Hunninghake, [61] W. Van Molle, C. Libert, How glucocorticoids control their own strength and the Macrophages survive hyperoxia via prolonged ERK activation due to phosphatase balance between pro- and anti-inflammatory mediators, Eur. J. Immunol. 35 down-regulation, J. Biol. Chem. 280 (2005) 26295–26302. (2005) 3396–3399. [48] Y. Ihara-Ohori, M. Nagano, S. Muto, H. Uchimiya, M. Kawai-Yamada, Cell death [62] T. Wu, C. Mohan, The AKT axis as a therapeutic target in autoimmune diseases, suppressor Arabidopsis bax inhibitor-1 is associated with calmodulin binding Endocr. Metab. Immune Disord. Drug Targets 9 (2009) 145–150. and ion homeostasis, Plant Physiol. 143 (2007) 650–660. [63] J.E. Niemela, L. Lu, T.A. Fleisher, J. Davis, I. Caminha, M. Natter, L.A. Beer, K.C. [49] N. Li, C. Wang, Y. Wu, X. Liu, X. Cao, Ca(2+)/calmodulin-dependent protein kinase II Dowdell, S. Pittaluga, M. Raffeld, V.K. Rao, J.B. Oliveira, Somatic KRAS mutations promotes cell cycle progression by directly activating MEK1 and subsequently associated with a human nonmalignant syndrome of autoimmunity and abnor- modulating p27 phosphorylation, J. Biol. Chem. 284 (2009) 3021–3027. mal leukocyte homeostasis, Blood. 117 (2010) 2883–2886. [50] M. Illario, A.L. Cavallo, K.U. Bayer, T. Di Matola, G. Fenzi, G. Rossi, M. Vitale, [64] X. Wu, J. Simpson, J.H. Hong, K.H. Kim, N.K. Thavarajah, P.H. Backx, B.G. Neel, T. Calcium/calmodulin-dependent protein kinase II binds to Raf-1 and modulates Araki, MEK–ERK pathway modulation ameliorates disease phenotypes in a integrin-stimulated ERK activation, J. Biol. Chem. 278 (2003) 45101–45108. mouse model of Noonan syndrome associated with the Raf1(L613V) mutation, [51] M. Le Bras, M.V. Clement, S. Pervaiz, C. Brenner, Reactive oxygen species and the J. Clin. Invest. 121 (2011) 1009–1025. mitochondrial signaling pathway of cell death, Histol. Histopathol. 20 (2005) [65] K. Liu, C. Liang, Z. Liang, K. Tus, E.K. Wakeland, Sle1ab mediates the aberrant 205–219. activation of STAT3 and Ras–ERK signaling pathways in B lymphocytes, J. Immunol. [52] S.W. Ryter, H.P. Kim, A. Hoetzel, J.W. Park, K. Nakahira, X. Wang, A.M. Choi, 174 (2005) 1630–1637. Mechanisms of cell death in oxidative stress, Antioxid. Redox Signal. 9 (2007) [66] . T. Kamata, J. Hussain, S. Giblett, R. Hayward, R. Marais, C. Pritchard, BRAF inactiva- 49–89. tion drives aneuploidy by deregulating CRAF, Cancer Res. 70 (2010) 8475–8486.