Portal de Periódicos da CAPES

A Novel Mitochondrial Ca2+-dependent Solute Carrier in the Liver Identified by mRNA Differential Display

2003; Elsevier BV; Volume: 278; Issue: 11 Linguagem: Inglês

10.1074/jbc.m208398200

1083-351X

Hirosato Mashima, Namiki Ueda, Hideki Ohno, Junko Suzuki, Hirohide Ohnishi, Hiroshi Yasuda, Tomohiro Tsuchida, Chiho Kanamaru, Noriko Makita, Taro Iiri, Masao Omata, Itaru Kojima,

Metabolism and Genetic Disorders

Pancreatic AR42J cells have the feature of pluripotency of the precursor cells of the gut endoderm. Dexamethasone converts them to exocrine cells or liver cells. Using mRNA differential display techniques, we have identified a novel Ca2+-dependent member of the mitochondrial solute carrier superfamily, which is expressed during the course of differentiation, and have designated it MCSC. The corresponding cDNA comprises an open reading frame of 1407 base pairs encoding a polypeptide of 469 amino acids. The carboxyl-terminal-half of MCSC has high similarity with other mitochondrial carriers, and the amino-terminal-half has three canonical elongation factor-hand motifs and has calcium binding capacity. The deduced amino acid sequence revealed 79.1% homology to the rabbit peroxisomal Ca2+-dependent member of the mitochondrial superfamily, but the subcellular localization of the protein was exclusively mitochondrial, not peroxisomal. Northern blot and Western blot analyses revealed its predominant expression in the liver and the skeletal muscle. In the liver, the expression level of MCSC was higher in the adult stage than in the fetal stage, and MCSC was highly up-regulated in dexamethasone-treated AR42J cells before the expression of albumin. Taken together, MCSC may play an important role in regulating the function of hepatocytes rather than in differentiation in vivo. Pancreatic AR42J cells have the feature of pluripotency of the precursor cells of the gut endoderm. Dexamethasone converts them to exocrine cells or liver cells. Using mRNA differential display techniques, we have identified a novel Ca2+-dependent member of the mitochondrial solute carrier superfamily, which is expressed during the course of differentiation, and have designated it MCSC. The corresponding cDNA comprises an open reading frame of 1407 base pairs encoding a polypeptide of 469 amino acids. The carboxyl-terminal-half of MCSC has high similarity with other mitochondrial carriers, and the amino-terminal-half has three canonical elongation factor-hand motifs and has calcium binding capacity. The deduced amino acid sequence revealed 79.1% homology to the rabbit peroxisomal Ca2+-dependent member of the mitochondrial superfamily, but the subcellular localization of the protein was exclusively mitochondrial, not peroxisomal. Northern blot and Western blot analyses revealed its predominant expression in the liver and the skeletal muscle. In the liver, the expression level of MCSC was higher in the adult stage than in the fetal stage, and MCSC was highly up-regulated in dexamethasone-treated AR42J cells before the expression of albumin. Taken together, MCSC may play an important role in regulating the function of hepatocytes rather than in differentiation in vivo. Both endocrine and exocrine cells of the pancreas arise from epithelial cells in the pancreatic duct (1Teitelman G. Lee J.K. Dev. Biol. 1987; 121: 454-466Google Scholar, 2Pictet R.L. Clark W.R. Williams R.H. Rutter W.J. Dev. Biol. 1972; 29: 436-467Google Scholar). Recently embryonic stem cells have been shown to differentiate into insulin-secreting structures similar to the pancreatic islet (3Lumelsky N. Blondel O. Laeng P. Velasco I. Ravin R. Mckay R. Science. 2001; 292: 1389-1394Google Scholar), and embryonic endodermal cells included a bipotential precursor population for pancreas and liver (4Deutsch G. Jung J. Zheng M. Lora J. Zaret K.S. Development. 2001; 128: 871-881Google Scholar). Pancreas originates from the dorsal and ventral regions of the foregut endoderm (5Slack J.M. Development. 1995; 121: 1569-1580Google Scholar), and liver develops from the ventral foregut endoderm adjacent to where the ventral pancreas emerges (6Zaret K.S. Mech. Dev. 2000; 92: 83-88Google Scholar). This proximity of budding sites for the liver and the ventral pancreas has led us to understand which molecular mechanisms control the differentiation of these organs. Recent genetic studies indicate that pancreatic development depends on an integrated network of distinct transcription factors operating at various levels. A mouse homeobox protein, insulin promoter factor-1 (IPF-1/PDX-1), is required for the development of the pancreas (7Jonsson J. Carlsson L. Edlund T. Edlund H. Nature. 1994; 371: 606-609Google Scholar). Islet-1, BETA2/NeuroD, Pax4, and Pax6 are necessary for the development and generation of mature islet cells (8Ahlgren U. Pfaff S.L. Jessell T.M. Edlund T. Edlund H. Nature. 1997; 385: 257-260Google Scholar, 9Naya F.J. Huang H.P. Qiu Y. Mutoh H. DeMayo F.J. Leiter A.B. Tsai M.J. Genes Dev. 1997; 11: 2323-2334Google Scholar, 10Sosa-Pineda B. Chowdhury K. Torres M. Oliver G. Gruss P. Nature. 1997; 386: 399-402Google Scholar, 11St-Onge L. Sosa-Pineda B. Chowdhury K. Mansouri A. Gruss P. Nature. 1997; 387: 406-409Google Scholar). The basic helix-loop-helix (bHLH) protein PTF1-p48 is essential for the formation of exocrine and the correct spatial organization of endocrine pancreas (12Krapp A. Knofler M. Ledermann B. Burki K. Berney C. Zoerkler N. Hagenbuchle O. Wellauer P.K. Genes Dev. 1998; 12: 3752-3763Google Scholar). In an endoderm explant assay, redundant fibroblast growth factor signaling from the cardiac mesoderm is necessary and sufficient to induce hepatogenesis within the ventral foregut endoderm (13Jung J. Zheng M. Goldfarb M. Zaret K.S. Science. 1999; 284: 1998-2003Google Scholar). Pancreatic AR42J cells are derived from a chemically induced pancreatic acinar cell tumor and have the feature of pluripotency of the common precursor cells of the pancreas (14Rosewicz S. Vogt D. Harth N. Grund C. Franke W.W. Ruppert S. Schweitzer E. Riecken E.O. Wiedenmann B. Eur. J. Cell Biol. 1992; 59: 80-91Google Scholar, 15Ohnishi H. Ohgushi N. Tanaka S. Nobusawa R. Mashima H. Mine T. Shimada O. Ishikawa H. Kojima I. J. Clin. Invest. 1995; 95: 2304-2314Google Scholar). We have shown that, when these cells were treated with activin A (Act) 1The abbreviations used are: Act, activin A; BTC, betacellulin; MCSC, mitochondrial Ca2+-dependent solute carrier; PBS, phosphate-buffered saline; UTR, untranslated region; dpc, day postcoitum; FITC, fluorescein isothiocyanate; GST, glutathioneS-transferase; Dx, dexamethasone and betacellulin (BTC) or hepatocyte growth factor, they differentiate into insulin-producing cells (16Mashima H. Ohnishi H. Wakabayashi K. Mine T. Miyagawa J. Hanafusa T. Seno M. Yamada H. Kojima I. J. Clin. Invest. 1996; 97: 1647-1654Google Scholar, 17Mashima H. Shibata H. Mine T. Kojima I. Endocrinology. 1996; 137: 3969-3976Google Scholar). When exposed to dexamethasone (Dx), they become more acinar-like cells (18Logsdon C.D. Mossner J. Williams J.A. Goldfine I.D. J. Cell Biol. 1985; 100: 1200-1208Google Scholar). It has been shown recently that when exposed to Dx for a long period, these cells synthesize albumin and have the character of liver cells (19Shen C.N. Slack J.M.W. Tosh D. Nat. Cell Biol. 2000; 2: 879-887Google Scholar). In this way, AR42J cells resemble the precursor cells in the gut, which differentiate into the pancreas and the liver and provide an excellent in vitro model system to study the differentiation of these organs. To clarify the molecular mechanisms of differentiation, we used the method of mRNA differential display and identified several genes that were up- or down-regulated during differentiation (20Mashima H. Yamada S. Tajima T. Seno M. Yamada H. Takeda J. Kojima I. Diabetes. 1999; 48: 304-309Google Scholar). Among them, we have focused on one gene that was highly up-regulated by Dx and characterized it. This gene was predominantly expressed in the liver and the skeletal muscle and was expressed at a higher level in the adult stage compared with the fetal stage. It is a novel member of the Ca2+-dependent mitochondrial carrier superfamily. Recombinant human Act was provided by Dr. Y. Eto of Central Research Laboratory, Ajinomoto Inc. (Kawasaki, Japan). Recombinant human BTC (21Seno M. Tada H. Kosaka M. Sasada R. Igarashi K. Shing Y. Folkman J. Ueda M. Yamada H. Growth Factors. 1996; 13: 181-191Google Scholar) was generously provided by Dr. M. Seno of Okayama University (Okayama, Japan). Hep3B cells were provided by the Cell Resource Center for Biomedical Research, Institute of Development, Aging, and Cancer, Tohoku University (Sendai, Japan). Hep3B cells were cultured in Dulbecco's modified Eagle's medium containing 20 mmol/liter HEPES/NaOH (pH 7.4), 5 mmol/liter NaHCO3, and 10% fetal bovine serum at 37° under a humidified condition of 95% air and 5% CO2. For staining, cells were grown on non-coated coverslips. 20 μg of total RNA extracted from AR42J cells were denatured and blotted onto a Hybond N+ nylon membrane (Amersham Biosciences). Human MTN Blot and Human Digestive System 12-Lane MTN Blot were purchased from Clontech (Palo Alto, CA). The blots were hybridized with 32P-labeled cDNA probe and washed for 30 min under high stringency conditions (0.1× standard saline citrate (SSC), 0.1% SDS) or low stringency conditions (2× SSC, 0.1% SDS) at 65° before exposure to x-ray film. From mRNA differential display, we obtained a 131-bp DNA fragment containing a putative poly(A) tail and a polyadenylation signal (20Mashima H. Yamada S. Tajima T. Seno M. Yamada H. Takeda J. Kojima I. Diabetes. 1999; 48: 304-309Google Scholar). cDNA was amplified using a gene-specific primer and a 5′-end-specific vector primer. Rat lung oligo(dT)-primed λZAPII cDNA library (Stratagene, La Jolla, CA) was used as a template. We first obtained a 956-bp cDNA fragment. A BLAST search using this fragment led to the identification of an expressed sequence tag from rat ovary mRNA (GenBankTM accession no. D86666), and we could extend the cDNA to 1128 bp. We went on gene walking by polymerase chain reaction (PCR) using a rat lung and brain cDNA library as a template. Finally we recovered the 3136-bp cDNA, which contained the 1407-bp open reading frame. The entire sequence of both strands was determined using an ABI PRISM BigDye Terminator Cycle Sequencing Ready reaction kit and an Applied Biosystems DNA sequencer 310 (Applied Biosystems, Cambridge, MA). To construct the MCSC expression vector, we amplified a fragment of MCSC (nucleotides 91–1507) by PCR using the sense primer (5′-ataagaatgcggccgctctcaccagcatgctctgcc-3′) and the antisense primer (5′-ccgctcgagtcatcaccgagactgcacgcccag-3′), followed by XhoI and NotI digestion. The fragment was subcloned into theXhoI/NotI sites of the vector pcDNA3.1/His (Invitrogen) and was verified by sequencing. pEGFP-Peroxi vector was purchased from Clontech. For transfection, FuGENE 6 transfection reagent (Roche Molecular Biochemicals) was used according to the manufacturer's instruction. Forty-eight hours after transfection, cells were fixed and stained. Protein extracts from rat liver, pancreas, stomach, colon, and skeletal muscle were prepared for immunoblotting as described previously (22Kanzaki M. Mashima H. Zhang Y.Q. Lu L. Shibata H. Kojima I. Nat. Cell Biol. 1999; 1: 165-170Google Scholar). All rat experiments were conducted in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and were approved by the University of Tokyo Institutional Animal Care and Use Committee. A polyclonal antibody against MCSC was raised by immunizing rabbits with peptide (C)KISEQQAEKILKSM (residues 113–126) conjugated with keyhole limpet hemocyanine. The antiserum was purified with a PD10 gel column (Amersham Biosciences). For Western blotting, membranes were blocked by incubation for 1 h with 10% blocking ace (Snow Brand, Japan) in phosphate-buffered saline (PBS). After blocking, the membranes were incubated with anti-MCSC antibody (1:100 dilution) for 1 h at room temperature, washed three times with PBS containing 0.1% Tween 20 for 10 min, and incubated for 30 min with horseradish peroxidase-conjugated goat anti-rabbit whole immunoglobulin (Nordic, 1:500 in PBS, 0.1% Tween 20). After washing, the enhanced chemiluminescence Western blotting detection reagent (Amersham Biosciences) was added, and the reaction was allowed to proceed according to the manufacturer's recommendations. Exposure to x-ray film was applied 1–3 min at room temperature. To remove the probe, membranes were incubated with the stripping buffer containing 62.5 mm Tris-HCl (pH 6.7) and 2% SDS for 30 min at 50°. Anti-actin-(1–19) antibody was purchased from Santa Cruz Biotechnology. For immunostaining, cells were grown on non-coated glass coverslips. Cells were fixed for 30 min in 3% paraformaldehyde in PBS, treated with 0.1% (v/v) Triton X-100 in PBS for 5 min, and incubated sequentially with Blocking Ace (Snow Brand, Tokyo, Japan), first antibody, and second antibody. Anti-Xpress Antibody (Invitrogen) was used to stain the exogenous MCSC. Second antibodies used in this study are indocarbocyanine (Cy3)-conjugated donkey anti-rabbit IgG, FITC-conjugated donkey anti-rabbit IgG, and FITC-conjugated anti-mouse IgG (Jackson Immuno Research Laboratories, West Grove, PA). MitoTracker Red CMXRos (Molecular Probes, Eugene, OR) and 4′,6-diamidino-2-phenylinodole,dihydrochloride (DAPI) (Molecular Probes, Eugene, OR) were used to stain mitochondria and nucleus respectively. The cells were examined under a light microscope (Axiophoto; Carl Zeiss, Inc., Thornwood, NY). Total RNA was extracted from rat liver by using TRIzol Reagent (Invitrogen). One-step RT-PCR was performed using SuperscriptTM One-Step RT-PCR with Platinum Taq (Invitrogen) according to the manufacturer's instruction. For semiquantitative RT-PCR, first-stranded cDNA was synthesized using SuperscriptTMfirst-stranded synthesis system for RT-PCR (Invitrogen) according to the manufacturer's instruction. Oligonucleotide primers used were 5′-tgtgctagtttcccaggaacc-3′ (nucleotides 10–30) and 5′-tcagcaggaccagcccagcg-3′ (nucleotides 1671–1652) for the whole coding region of MCSC (1662-bp PCR product); 5′-tgtgagcatcagctacgtgg-3′ (nucleotides 1447–1466) and 5′-ggttccaggttctagcactag-3′ (nucleotides 2167–2147) for the semiquantitative RT-PCR of MCSC (721 bp PCR product); and 5′-tgagagggaaatcgtgcgtg-3′ (nucleotides 612–631) and 5′-gatccacatctgctggaaggtg-3′ (nucleotides 1071–1051) for β-actin (GenBankTM accession no. V01217, 460-bp PCR product). PCR cycles for semiquantitative RT-PCR were 25 for MCSC and 20 for β-actin. To construct the amino-terminal fragment of MCSC, we amplified a fragment of MCSC (residues 1–178) by PCR using the sense primer (5′-ccgctcgagcatgctctgcctgtgcctgtatg-3′) and the antisense primer (5′-ccgctcgagcactgtgaactcatctgggac-3′), followed byXhoI digestion. The fragment was subcloned into theXhoI site of the vector, pGEX-4T-3 (Amersham Biosciences) and was verified by sequencing. Escherichia coli strain BL21 was transfected with the final construct or pGEX-4T-3. The purification of the GST fusion protein was essentially done according to the recommendation of the manufacturer of glutathione-Sepharose 4B (Amersham Biosciences). Proteins were resolved in SDS-PAGE (10%) and transferred to a nitrocellulose membrane. 45Ca2+ overlay was then performed essentially as described (23Maruyama K. Mikawa T. Ebashi S. J. Biochem. (Tokyo). 1984; 95: 511-519Google Scholar). The membrane was washed in a solution containing 60 mmol/liter KCl, 5 mmol/liter MgCl2, and 10 mmol/liter imidazole-HCl (pH 6.8) for 1 h with three changes of the buffer. Then the membrane was incubated in the same buffer containing 10 μCi of 45Ca2+/ml for 10 min. After incubation, the membrane was rinsed twice with 40% ethanol for 5 min. Excess water was absorbed between two sheets of Whatman no. 1 filter paper, and the membrane was dried at room temperature and exposed to x-ray film at −80°. In mRNA differential display, we identified several novel genes up-regulated by more than 10-fold when AR42J cells were exposed to Dx (20Mashima H. Yamada S. Tajima T. Seno M. Yamada H. Takeda J. Kojima I. Diabetes. 1999; 48: 304-309Google Scholar). We only had small fragments of the 3′-untranslated region (UTR) with which to obtain the full-length cDNA of these genes. During the cloning experiments, some genes were recorded in the GenBankTM data base as glutaredoxin, adenine nucleotide translocator, and so on. We picked up one gene, which was highly up-regulated when exposed to Dx, and the size of mRNA was estimated to be about 3.5 kbp according to Northern blot analysis (Fig. 1). First we obtained the 131-bp 3′-fragment containing a putative poly(A) tail and a polyadenylation signal. Using the technique of gene walking by PCR and a BLAST search, we obtained the 3136-bp cDNA, which comprises an open reading frame of 1407 bp, a 100-bp 5′-UTR, and a 1638-bp 3′-UTR (Fig. 2 A).— There may be additional regions of the 5′-UTR to be cloned; however, there is an in-frame stop codon in the 5′-UTR of our sequence. The sequence just before the initiation codon is in good agreement with the Kozak sequence (24Kozak M. Nucleic Acids Res. 1987; 15: 8125-8132Google Scholar). A consensus polyadenylation signal (AATAAA), nucleotides 3115–3120, was located 15 bases before the poly(A) tail. The open reading frame encodes a predicted polypeptide of 469 amino acids with a calculated molecular mass of 52.7 kDa.Figure 2cDNA and protein sequences of MCSC.A, cDNA and protein sequences of MCSC are shown. The putative open reading frame is indicated in uppercase letters, and 5′-UTR and 3′-UTR are indicated in lowercase letters. The in-frame stop codon in 5′-UTR and the possible polyadenylation signal are underlined. B, comparison of the predicted amino acid sequence of MCSC and Efinal. Anasterisk indicates identity, a colon indicates conservatively substituted, and a period indicates similarity. Three well conserved EF-hands are indicated withhorizontal lines. Amino acid sequences were aligned with ClustalW (version 1.8). C, alignment of MCSC with other rat mitochondrial carrier sequences. The sequences shown are solute carrier family 25, member 5 (adenine nucleotide translocator 2, fibroblast form; Slc25a5: D12771); solute carrier family 25, member 4 (adenine nucleotide translocator; Slc25a4: D12770); dicarboxylate carrier (CAA11278); solute carrier family 25, member 1 (citrate transporter: NM017307); and oxodicarboxylate carrier (odc gene: AJ289714). Six predicted transmembrane domains are indicated by horizontal lines; their locations are based on human ADP/ATP translocase (AAC2 isoform; M57424) transmembrane domain positions. Amino acid sequences were aligned with ClustalW (version 1.8). D, phylogenetic tree of MCSC, Efinal, and the members of rat mitochondrial transporter superfamily. The scale shows the evolutionary distances calculated.View Large Image Figure ViewerDownload (PPT)Figure 2cDNA and protein sequences of MCSC.A, cDNA and protein sequences of MCSC are shown. The putative open reading frame is indicated in uppercase letters, and 5′-UTR and 3′-UTR are indicated in lowercase letters. The in-frame stop codon in 5′-UTR and the possible polyadenylation signal are underlined. B, comparison of the predicted amino acid sequence of MCSC and Efinal. Anasterisk indicates identity, a colon indicates conservatively substituted, and a period indicates similarity. Three well conserved EF-hands are indicated withhorizontal lines. Amino acid sequences were aligned with ClustalW (version 1.8). C, alignment of MCSC with other rat mitochondrial carrier sequences. The sequences shown are solute carrier family 25, member 5 (adenine nucleotide translocator 2, fibroblast form; Slc25a5: D12771); solute carrier family 25, member 4 (adenine nucleotide translocator; Slc25a4: D12770); dicarboxylate carrier (CAA11278); solute carrier family 25, member 1 (citrate transporter: NM017307); and oxodicarboxylate carrier (odc gene: AJ289714). Six predicted transmembrane domains are indicated by horizontal lines; their locations are based on human ADP/ATP translocase (AAC2 isoform; M57424) transmembrane domain positions. Amino acid sequences were aligned with ClustalW (version 1.8). D, phylogenetic tree of MCSC, Efinal, and the members of rat mitochondrial transporter superfamily. The scale shows the evolutionary distances calculated.View Large Image Figure ViewerDownload (PPT) Searching the protein data base with the deduced amino acid sequence revealed conservation with the rabbit peroxisomal Ca2+-dependent solute carrier (Efinal, GenBankTM accession no. T50686): 61.6% identity and 79.1% similarity (Fig. 2 B), and Homo sapienscalcium-binding transporter (GenBankTM accession no.AF123303): 58.2% identity and 73.0% similarity. However, the first 51 amino acids have no homology with other proteins in the data base. The amino-terminal-half contains three well-conserved Ca2+-elongation factor (EF)-hand binding loops (residues 60–72, 91–103, and 127–139). These EF-hand domains are also conserved in Efinal (24Kozak M. Nucleic Acids Res. 1987; 15: 8125-8132Google Scholar). The carboxyl-terminal-half has substantial similarity with proteins of the mitochondrial solute carrier family (16.4–21.9% identity) (Fig. 2 C). The hydropathic profile of this region revealed six transmembrane domains, similar to other mitochondrial carriers (data not shown). We designated this protein as MCSC. Fig. 2 D depicts the phylogenetic tree of MCSC, Efinal, and other rat mitochondrial carriers. Recently human KIAA1896 protein (GenBankTM accession no. AB067483) has been reported in the data base (26Kikuno R. Nagase T. Waki M. Ohara O. Nucleic Acids Res. 2002; 30: 166-168Google Scholar). Residues 137–568 of this protein have 93.7% identity and 95.8% similarity with residues 50–469 of MCSC. With its strong homology, the KIAA1896 protein may be a human ortholog of MCSC, but the first 136 amino acids has no relationships with the first 49 amino acids of MCSC (Fig. 3). There may be an alternative splicing of MCSC but only one transcript could be seen on the Northern blot (Figs. 4 and 5). KIAA1896 may be an isoform encoded by different genes. We have to exclude the possibility of the merger of two cDNAs in MCSC cDNA. Using a rat liver total RNA as a template, we can amplify the whole coding region by RT-PCR (Fig.6). Recently mouse clone IMAGE: 4239441 (GenBankTM accession no. BC019978) and IMAGE: 5098924 (GenBankTM accession no. BC022114) have been recorded in the data base. At the nucleotide level these have 93.7% identity through nucleotides 61–2401 of MCSC and 99.5% identity at the protein level. These genes were predicted to be the mouse gene of MCSC but are 103 amino acids shorter than MCSC. This may result from the misidentification of the initiation codon. These are translated from the second ATG codon of their sequences. When translated from the first ATG codon, 103 amino acids are added, and there is a 99.1% identity over the whole region at the protein level (Fig. 3). During the preparation of this article, another version of the KIAA1896 protein was reported by NCBI Annotation Project (GenBankTMaccession no. XP 027668). This predicted KIAA1896 protein sequence was almost identical to that of MCSC over the whole coding region: 97.0% identity and 98.7% similarity. This gene might be the real human ortholog of MCSC. Some other proteins with high similarity to MCSC were reported in the data base: GenBankTM accession no.AAH05163, 95.2% similarity (H. sapiens, 311 amino acids);AAH37109, 88.1% similarity (mouse, 514 amino acids); BAB70825, 71.4% similarity (H. sapiens, 384 amino acids); XP 131097, 59.4% similarity (mouse, 475 amino acids, may be a mouse ortholog of Efinal). These proteins have high similarities at the carboxyl terminus of MCSC but have low similarities at the amino terminus. In particular, there are no similarities between the first 51 amino acids of MCSC and these proteins.Figure 4Northern blot analysis of MCSC expression in various tissues. mRNA (2 μg) isolated from various tissues was blotted onto a nylon membrane, and Northern blotting was performed using 32P-labeled MCSC cDNA. The blot was reprobed with32P-labeled β-actin (lower panel).View Large Image Figure ViewerDownload (PPT)Figure 5Northern blot analysis of MCSC expression in the digestive tract. mRNA (2 μg) isolated from various tissues was blotted onto a nylon membrane, and Northern blotting was performed using 32P-labeled MCSC cDNA. In the transverse colon lane, we could not remove the strong stain. The blot was reprobed with 32P-labeled β-actin (lower panel).View Large Image Figure ViewerDownload (PPT)Figure 6RT-PCR of the whole coding region of MCSC. One-step RT-PCR was performed using rat liver total RNA (1 μg) as a template. Samples with (lane 1) or without (lane 2) reverse transcriptase were loaded. The right lane is a size marker of HindIII-digested λ DNA. The band was confirmed by sequencing.View Large Image Figure ViewerDownload (PPT) Northern blot analysis was carried out to examine the tissue distribution of MCSC. Samples of poly(A) mRNA isolated from human tissues were analyzed with labeled probe specific for MCSC. MCSC was expressed predominantly in the liver and the skeletal muscle as a single transcript (Fig. 4). With a long exposure time, weak levels of MCSC expression were detected in all the tissues examined (data not shown). Efinal was highly expressed in the colon, followed by small intestine and kidney. In the small intestine, the expression of Efinal was the lowest in duodenum and increased toward the distal part of the intestine (25Weber F.E. Minestrini G. Dyer J.H. Werder M. Boffelli D. Compassi S. Wehrli E. Thomas R.M. Schulthess G. Hauser H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 8509-8514Google Scholar). We then examined the expression level of MCSC in digestive tract using the Human Digestive System 12-Lane MTN BLOT. Very weak levels of expression, compared with the liver, were detected through the esophagus to the rectum (Fig.5). Western blot analysis was then carried out to clarify endogenous MCSC protein expression in various rat tissues. The expression with a molecular mass of 53 kDa was detected in skeletal muscle and liver, consistent with Northern blot analysis (Fig. 7). To investigate the subcellular localization of MCSC protein, we performed immunofluorescence study using a specific antibody to MCSC. In the cytoplasm of Hep3B cells, both punctate and filamentous patterns were observed, consistent with MCSC protein being present in intracellular organelles (Fig. 8 A). We could not eliminate nuclear staining, but when we expressed the MCSC protein ectopically using the expression vector MCSC-pcDNA3.1/His, we could see immunostaining only in the cytoplasm (Fig. 8 B). So we believe nuclear staining is nonspecific. Omission of the primary antibody resulted in the complete absence of immunostaining (data not shown). To confirm that MCSC-positive cytoplasmic organelles were mitochondrial, we treated Hep3B cells with mitochondrial-specific fluorescent dye MitoTracker Red CMXRos to label mitochondria in the first antibody incubation step. As shown in Fig. 8 C, the MCSC staining pattern overlapped with that of mitochondria, demonstrating that MCSC is targeted to mitochondria. Efinal is mainly localized to peroxisomes, and some are expressed in mitochondria (25Weber F.E. Minestrini G. Dyer J.H. Werder M. Boffelli D. Compassi S. Wehrli E. Thomas R.M. Schulthess G. Hauser H. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 8509-8514Google Scholar). We then examined whether MCSC is targeted to peroxisomes. When the pEGFP-Peroxi vector is transfected to cells, it specifically targets to peroxisomal membranes. As shown in Fig.8 D, the staining pattern of MCSC and peroxisomes are not superimposable, indicating that MCSC is not expressed in peroxisomes. In rat liver strong immunoreactivity against MCSC was observed in the cytoplasm as punctate and round patterns compatible with its expression in mitochondria of hepatocytes (Fig.8 E). Since the amino terminus of MCSC contains three canonical EF-hands (residues 60–72, 91–103, and 127–139) that are predicted to bind calcium, we expressed a GST-tagged truncated version of MCSC (residues 1–178) in E. coli and purified it to homogeneity (Fig.9 A). The calcium binding ability of the polypeptide was demonstrated with a Ca2+-overlay method. As shown in Fig. 9 B, a truncated version of MCSC exhibits an affinity to calcium even in the SDS-denatured state. MCSC was up-regulated when AR42J cells were incubated with Dx and predominantly expressed in liver and skeletal muscle. Because MCSC might be involved in the differentiation of hepatocytes, we compared the expression level in the fetal stage (18.5-day postcoitum (dpc)), the neonatal stage (2-day-old), and the adult stage (9-week-old). As shown in Fig.10 A, the expression level of MCSC was the greatest in the adult stage. When we compared the mRNA level, the expression level of MCSC was higher in the adult stage (9-week-old) than in the fetal stage (18.5 dpc) (Fig. 10, Band C). In the present study, we have identified a novel cDNA, MCSC, encoding a protein with 469 amino acids belonging to a Ca2+-dependent member of the mitochondrial transporter superfamily. Mitochondrial carriers have a tripartite structure, made up of related sequences about 100 amino acids in length (27Palmieri F. FEBS Lett. 1994; 346: 48-54Google Scholar). Each repetitive element contains two hydrophobic stretches separated by an extensive hydrophilic regions (I-II, III-IV, and V-VI) (27Palmieri F. FEBS Lett. 1994; 346: 48-54Google Scholar). The carboxyl-terminal-half of MCSC is compatible with these criteria and has a homology with other mitochondrial carrier proteins (Fig. 2 C). An intriguing feature of MCSC is the presence of three EF-hand-like domains in the amino terminus (Fig. 2 B). The structural similarity by itself does not prove that EF-hand like domains of MCSC bind calcium, but the strong homology to Efinal, and the Ca2+ binding capacity of Efinal strongly suggest that EF-hand like domains of MCSC bind calcium. Another Ca2+binding member of mitochondrial carrier superfamily, Aralar (GenBankTM accession no. Y14494) was reported and was present in human muscle and brain (28Arco A. Satrustegui J. J. Biol. Chem. 1998; 273: 23327-23334Google Scholar). There are four EF-hand-like domains in the amino-terminal-half and a tripartite structure in the carboxyl-terminal-half of Aralar. There is a 16.8% identity between MCSC and Aralar, and the amino-terminal-half of Aralar also has calcium binding activity (28Arco A. Satrustegui J. J. Biol. Chem. 1998; 273: 23327-23334Google Scholar). We examined the Ca2+ binding properties of MCSC with a truncated protein, and the amino terminus of MCSC clearly showed Ca2+ binding activity (Fig. 9). Mitochondrial carriers are encoded by nuclear genes and have to be imported into the mitochondrial membranes. Like most mitochondrial carriers except phosphate and citrate carriers in mammals (29Runswick M.J. Powell S.J. Nyren P. Walker J.E. EMBO J. 1987; 6: 1367-1373Google Scholar, 30Runswick M.J. Walker J.E. Bisaccia F. Iacobazzi V. Palmieri F. Biochemistry. 1990; 29: 11033-11040Google Scholar), MCSC has no obvious amino-terminal mitochondrial import sequence. The carboxyl-terminal Ser-Lys-Leu (SKL) sequence was the first recognized peroxisomal-targeting sequence (31Gould S.J. Keller G.A. Hosken N. Wilkinson J. Subramani S. J. Cell Biol. 1989; 108: 1657-1664Google Scholar). It is not present in MCSC. But without it, Efinal is mainly targeted to the peroxisomal membrane (28Arco A. Satrustegui J. J. Biol. Chem. 1998; 273: 23327-23334Google Scholar). Localization studies presented here show clearly that MCSC is expressed in mitochondria, not in peroxisome (Fig. 8, C andD). When expressed ectopically, MCSC was also located in mitochondria (Fig. 8 B). MCSC has three EF-hand-like domains in the amino terminus, and intracellular Ca2+ concentration may be related to the localization of MCSC. In rat hepatocytes, vasopressin, endothelin, and angiotensin II elevate intracellular Ca2+ concentration. When we add these secretagogues to the culture medium in Hep3B cells, there are no significant changes in the MCSC staining pattern (data not shown). AR42J cells were originally derived from a rat pancreatic acinar cell tumor. They possess both exocrine and neuroendocrine properties (14Rosewicz S. Vogt D. Harth N. Grund C. Franke W.W. Ruppert S. Schweitzer E. Riecken E.O. Wiedenmann B. Eur. J. Cell Biol. 1992; 59: 80-91Google Scholar). Synthetic glucocorticoid, Dx, can increase the expression of amylase in AR42J cells and enhance the differentiation of AR42J cells toward the exocrine phenotype (18Logsdon C.D. Mossner J. Williams J.A. Goldfine I.D. J. Cell Biol. 1985; 100: 1200-1208Google Scholar). When a subclone of AR42J cells, AR42J-B13, was incubated for more than 9 days with dexamethasone, the expression of amylase was almost gone, and the expression of albumin was observed, indicating that the cells converted into hepatocytes (19Shen C.N. Slack J.M.W. Tosh D. Nat. Cell Biol. 2000; 2: 879-887Google Scholar). Dx also induced the transdifferentiation of organ cultures of pancreatic buds from mouse embryo into hepatocytes (19Shen C.N. Slack J.M.W. Tosh D. Nat. Cell Biol. 2000; 2: 879-887Google Scholar). The liver and the pancreas originate from neighboring regions of the foregut endoderm (5Slack J.M. Development. 1995; 121: 1569-1580Google Scholar, 6Zaret K.S. Mech. Dev. 2000; 92: 83-88Google Scholar), and some molecular and cellular mechanisms are regulating the development of these two organs. MCSC was highly up-regulated when AR42J cells were incubated with Dx for 3 days (Fig. 1), and MCSC was predominantly expressed in the liver and skeletal muscle (Fig. 4). We therefore suggest that MCSC was up-regulated during the differentiation of AR42J cells toward hepatocytes rather than exocrine pancreas. Glucose-6-phosphatase is normally found in the liver as well as in the β-cells of pancreas and is one of the earliest differentiation markers. Glucose-6-phosphatase-positive cells could be seen after 3 days of addition of Dx in AR42J-B13 cells (19Shen C.N. Slack J.M.W. Tosh D. Nat. Cell Biol. 2000; 2: 879-887Google Scholar) when MCSC was up-regulated. In contrast, albumin, a liver-specific protein, started to appear after 9 days (19Shen C.N. Slack J.M.W. Tosh D. Nat. Cell Biol. 2000; 2: 879-887Google Scholar). Taken together, MCSC may play a role during the differentiation toward hepatocytes. Fibroblast growth factors have been implicated in the primary induction of the liver (6Zaret K.S. Mech. Dev. 2000; 92: 83-88Google Scholar). It remains to be seen whether fibroblast growth factors regulate the expression of MCSC. Western blotting and RT-PCR study showed that the expression level of MCSC was higher in the adult stage compared with the fetal stage and the neonatal stage (Fig. 10) indicating that MCSC may play an important role in regulating the function of hepatocytes rather than the differentiation in vivo. In the study of mRNA differential display, we have identified other mitochondrial carriers up- or down-regulated over the course of differentiation. Adenine nucleotide translocator was up-regulated when incubated with Dx, and citrate transporter-like protein was up-regulated when incubated with Act + BTC. 2H. Mashima, unpublished data. Similar to the importance of the ATP/ADP ratio in insulin secretion of pancreatic β-cells, the cellular environment resulting from the functions of these mitochondrial proteins including MCSC may be important in regulating the function and the differentiation of pancreas and liver. MCSC may act as a calcium binding/transducing protein by the transport of a solute. Further studies are needed to confirm our speculations. We thank Dr. Yositaka Konda of the Department of Gastroenterology, Kyoto University for helpful discussions.