Molecular characterization of mitocalcin, a novel mitochondrial Ca2+-binding protein with EF-hand and coiled-coil domains

Authors


Address correspondence and reprint requests to Yasuhiro Tomooka, Department of Biological Science and Technology and Tissue Engineering Research Center, Tokyo University of Science, 2641 Yamazaki, Noda, Chiba 278-8510, Japan. E-mail: tomoylab@rs.noda.tus.ac.jp

Abstract

Here we have identified and characterized a novel mitochondrial Ca2+-binding protein, mitocalcin. Western blot analysis demonstrated that mitocalcin was widely expressed in mouse tissues. The expression in brain was increased during post-natal to adult development. Further analyses were carried out in newly established neural cell lines. The protein was expressed specifically in neurons but not in glial cells. Double-labeling studies revealed that mitocalcin was colocalized with mitochondria in neurons differentiated from 2Y-3t cells. In addition, mitocalcin was enriched in the mitochondrial fraction purified from the cells. Immunohistochemical studies on mouse cerebellum revealed that the expression pattern of mitocalcin in glomeruli of the internal granular and molecular layers was well overlapped by the distribution pattern of mitochondria. Immunogold electron microscopy showed that mitocalcin was associated with mitochondrial inner membrane. Overexpression of mitocalcin in 2Y-3t cells resulted in neurite extension. Inhibition of the expression in 2Y-3t cells caused suppression of neurite outgrowth and then cell death. These findings suggest that mitocalcin may play roles in neuronal differentiation and function through the control of mitochondrial function.

Abbreviations used
COX I

cytochrome oxidase subunit I

FCS

fetal calf serum

HA

hemagglutinin

PBS

phosphate-buffered saline

SF + 4F

serum-free medium plus 4 factors

TBS-T

Tris-buffered saline, 0.05% Tween 20

Neurons in the mammalian CNS are highly specialized in morphology and function and are generated from neuroepithelial cells of a neural tube (Jacobson 1991). The identity of specific neurons is established through a sequence of steps, including neurogenesis, cell migration, laminar positioning, neurite extension and synaptogenesis. Each of the steps in the sequence has been extensively studied, although limited information is as yet available on the underlying molecular processes.

Mitochondria are essential organelles in eukaryotic cells and their major functions are generating ATP and regulating cellular Ca2+ homeostasis (McCormack and Denton 1993; Simpson and Russell 1998). Some studies report that mitochondria are involved in neurogenesis, neurite outgrowth and maintenance of neuronal functions. For example, when neural stem cells generate neurons in mouse forebrain, there is a decrease of the mitochondrial membrane potential in neural stem cells (Ohsawa et al. 2004), indicating that mitochondria may be related to neurogenesis in the CNS. Examination of organellar distribution in embryonic hippocampal neurons also reveals a concentration of mitochondria at the base of a presumptive axon in the very early stages of axogenesis (Bradke and Dotti 1997; Mattson and Partin 1999; Ruthel and Hollenbeck 2003), suggesting a possible role of mitochondria in the process of axogenesis. Furthermore, neurite outgrowth is influenced greatly by cytoplasmic free Ca2+ levels and energy availability. Ca2+ regulates cytoskeletal dynamics and membrane trafficking (Mattson 1992; De Camilli and Takei 1996), while ATP is required for running motor proteins and for phosphorylation reactions (Bridgman et al. 1994; Desai et al. 1997). In addition, mitochondrial Ca2+ regulation has been proposed to play roles in a variety of physiological events including synaptic plasticity (Tang and Zucker 1997; Levy et al. 2003) and excitotoxic neuronal death (Nicholls and Budd 1998). Thus, mitochondria play important roles in neuronal differentiation and function and the dysfunction seems to be one of the causes of neurodegenerative diseases (Wallace 1999; Schapira 2000; Schon 2000).

We have recently reported developmental characteristics of a clonal neural progenitor cell line 2Y-3t (Tominaga et al. 2005). The cells cultured in serum-containing medium proliferate and express marker proteins of neural progenitor/stem cells (undifferentiated 2Y-3t cells). When the cells are cultured in serum-free medium, they extend long cytoplasmic processes and expression of neuronal marker proteins is increased in concert with the morphological differentiation (differentiated 2Y-3t cells). We have isolated genes differentially expressed in differentiated 2Y-3t cells with a subtraction method (Tominaga and Tomooka 2002). In the present study, we describe one of the genes encoding a novel mitochondrial Ca2+-binding protein, mitocalcin, involved in neuronal differentiation and function.

Materials and methods

Animals

CD-1 mice (Charles River Japan, Yokohama, Japan) were maintained in the experimental animal facility of Tokyo University of Science. They were kept under a 12/12 h light/dark cycle at 22–24°C. Standard laboratory feed (MR standard, Nousan Ltd, Yokohama, Japan) and tap water were given ad libitum. Mouse care and handling conformed to the NIH guidelines for animal research. The experimental protocols were approved by the Tokyo University of Science Animal Care and Use Committee.

Cell culture

A clonal cell line 2Y-3t was established from a cerebellum of an adult p53-deficient mouse (Minakawa et al. 1998) and was characterized as neural progenitor cells (Tominaga et al. 2005). The cells were maintained in medium [1 : 1 mixture of Dulbecco's modified Eagle's medium and Ham's F-12 without phenol red, Sigma Chemical Company, St Louis, MO, USA] containing heat-inactivated fetal calf serum (FCS) at 10% (Gibco-BRL, Burlington, Ontario, Canada) supplemented with 31 µg/mL penicillin (Sigma Chemical Company), 50 µg/mL streptomycin (Sigma Chemical Company), 10% FCS medium at 37°C, 5% CO2. The cells were also cultured on poly-d-lysine- (Sigma Chemical Company) coated dishes or coverslips in Dulbecco's modified Eagle's medium and Ham's F-12 supplemented with 10 µg/mL insulin, 10 µg/mL transferrin, 10−8 m sodium selenite and 10 ng/mL cholera toxin (all Sigma Chemical Company) (SF + 4F medium) at 37°C, 5% CO2.

Clonal glial cell lines FBD-102b and FBD-104 were established from brains of p53-deficient mice at embryonic day 13 (Tomooka and Aizawa 1998; Horiuchi and Tomooka, unpublished observations). FBD-102b cells were cultured in Dulbecco's modified Eagle's medium and Ham's F-12 containing heat-inactivated FCS at 0.5% (0.5% FCS medium) and FBD-104 cells were cultured in 10% fetal calf serum plus 4 factors (10% FCS + 4F) medium.

Preparation of anti-mitocalcin antibody

Mouse mitocalcin cDNA was obtained by a subtractive screening as described by Tominaga and Tomooka (2002). For the construction of His-tagged mitocalcin protein, a cDNA encoding a part of the protein (amino acids 160–240, showed by underlining in Fig. 1a) was subcloned into a pTrcHis B vector (Invitrogen, Carlsbad, CA, USA). The fusion protein was purified by immobilized Ni2+ absorption chromatography. Rabbits were injected with the recombinant protein, initial immunization was performed with Freund's complete adjuvant (Wako Pure Chemical Industries, Osaka, Japan) and boosters were with Freund's incomplete adjuvant (Wako Pure Chemical Industries). Specificity of the antiserum was tested by western blot analysis.

Figure 1.

Primary structure of mouse mitocalcin. (a) Schematic representation of the domain structure of mitocalcin. Mitocalcin had two EF-hand domains (amino acids 95–123 and 131–159) in the center and a coiled-coil domain (amino acids 203–237) at the C-terminal end. (b) Alignment of mouse mitocalcin and putative human homolog (Accession no. NM025202). The amino acid residues shared by the homolog are shown by black boxes; missing amino acids are indicated by dashes; residues corresponding to EF-hand domains are shown by boxes; residues corresponding to coiled-coil domain are underlined and amino acid residues of mouse mitocalcin are numbered on the right.

Western blot analysis

The protocol for western blot analysis has been described previously (Ohkawara et al. 2003). Membranes were blocked by incubating them in 10% non-fat dried milk in Tris-buffered saline, 0.05% Tween 20 (TBS-T) and immersed in working dilutions of primary antibodies in TBS-T, 5% milk. After three washes in TBS-T containing 5% milk, membranes were incubated with horseradish peroxidase-conjugated secondary antibody (Amersham Pharmacia, Buckinghamshire, UK) for 1 h. They were then washed twice with TBS-T containing 5% milk, once with TBS-T and incubated in a 1 : 1 mixture of enhanced chemiluminescence reaction solution (Amersham Pharmacia) at 20–25°C. Autoradiographs were prepared by exposing the membranes to X-ray film (Kodak Japan, Tokyo, Japan).

Subcellular fractionation

Subcellular fractions for western blot analysis were prepared using the Mitochondria Isolation Kit (Pierce, Rockford, IL, USA) according to the manufacturer's instructions. Purity of the mitochondria-enriched fraction was confirmed using anti-cytochrome oxidase subunit I (COX I) (Molecular Probes, Eugene, OR, USA) antibody. To assess the functional or morphological integrity, the isolated mitochondria were examined by western blot analysis for cytochrome c (Sigma Chemical Company) and Voltage-Dependent Anion Channel (VDAC) (Sigma Chemical Company) and by an electron microscope.

Immunocytochemistry

Cells were fixed with 4% paraformaldehyde in 0.1 m phosphate buffer (pH 7.4) for 10 min and permeabilized in 0.2% Triton X-100/phosphate-buffered saline (PBS) for 5 min at room temperature. Cells were blocked in 5% normal goat serum/PBS (blocking solution) for 1 h at room temperature. Cells were then overlaid with primary antibodies at working dilutions and incubated for 1 h at room temperature. The primary antibodies were rabbit anti-mitocalcin, mouse monoclonal anti-GM130 (BD-Transduction Laboratories, San Jose, CA, USA), anti-COX I, anti-β-tubulin III (TuJ-1, Babco, Richmond, CA, USA), anti-glial fibrillary acidic protein (Sigma Chemical Company), rat monoclonal anti-GRP94 (Abcam, Cambridgeshire, UK) and anti-myelin basic protein (Serotec, Oxford, UK) antibodies. To visualize actin filaments, cells were stained with rhodamine-phalloidin (Molecular Probes). After washing with PBS, cells were incubated with secondary antibodies at room temperature for 1 h. The secondary antibodies were FITC- or TRITC-conjugated goat anti-mouse, anti-rabbit or anti-rat IgG (Jackson Immunoresearch, West Grove, PA, USA) antibodies. Fluorescence specimens were viewed with a fluorescence microscope (Axiovert 200, Carl Zeiss, Oberkochen, Germany) or a confocal laser-scanning microscope (LSM510, Carl Zeiss).

Immunohistochemistry

Mouse brains were perfused with 4% paraformaldehyde in 0.1 m phosphate buffer under anesthesia with ether and then cut into small pieces. These samples were immersed in the same fixative for 1 h. After washing with PBS, tissues were immersed successively in PBS solutions containing 10, 15 and 20% sucrose. After tissues were embedded in OCT compound and frozen, cryosections (thickness 4–6 µm) were cut with a Jung Frigocut 2800E (Leica, Wetzlar, Germany) and then mounted on silane-coated glass slides. The cryosections were rinsed with PBS and permeabilized in 0.2% Triton X-100/PBS for 5 min. The sections were incubated for 1 h at room temperature with rabbit anti-mitocalcin and mouse monoclonal anti-COX I antibodies. The sections were then incubated for 1 h at room temperature with secondary antibodies as described above.

Immunogold labeling of ultrathin cryosections

The protocol for immunogold labeling analysis has been described previously (Ichimura et al. 2003). Ultrathin cryosections were incubated with rabbit anti-mitocalcin antibody at 4°C overnight. They were then incubated with anti-rabbit IgG (British BioCell, Essex, UK) coupled to 10-nm colloidal gold for 1 h at room temperature. After immunostaining, they were fixed with 2.5% glutaraldehyde in phosphate buffer. The sections were then contrasted with 2% uranyl acetate for 20 min and absorption-stained with 3% polyvinyl alcohol containing 0.2% uranyl acetate for 10 min. As a negative control experiment, the primary antibodies were either omitted or replaced with normal rabbit IgG.

The quantitative analysis of the gold particle distribution in mitochondria was estimated by counting the total number of gold particles in mitochondria and the number of gold particles in mitochondrial compartments. The quantitative result was expressed as mean ± SEM values.

Observation of mitochondrial Ca2+ in 2Y-3t cells

Cells were incubated in PBS containing 5 µm of the acetoxymethylester form of magfura-2 (Molecular Probes), 0.5% dimethylsulfoxide and 5 mg/mL bovine serum albumin for 15 min at 37°C. The cells were then washed with PBS and observed with an AS-MDW (Leica) for baseline measurement of cytosolic free calcium. The cells were then incubated in PBS (without Ca2+ and Mg2+) containing 750 nm carbonylcyanide-p-(trifluoromethoxy)-phenylhydrazone (Sigma Chemical Company) at 20–25°C and the intensity of cytosolic free calcium was observed with an AS-MDW.

Transfection of plasmid into 2Y-3t cells

Sense and antisense mitocalcin cDNAs were inserted into the mammalian expression vector pcDNA3.1/Hygro (Invitrogen). 2Y-3t cells were grown in 60-mm dishes in 10% FCS medium until reaching 60–70% confluence. Dulbecco's modified Eagle's medium and Ham's F-12 (100 µL) was mixed with Lipofectin® Reagent (10 µL; Invitrogen) and with pcDNA3.1 sense or antisense mitocalcin vectors and incubated at room temperature for 30 min. The transfection solution was added to each dish. After 24 h, the medium was removed and replaced with 10% FCS medium. Approximately 2 weeks after transfection, hygromycin-resistant colonies were selected and expanded in 10% FCS medium supplemented with 1 mg/mL hygromycin (Sigma Chemical Company). Several clones were characterized with western blot analysis as described above.

A specific mouse mitocalcin small interfering RNA (siRNA) (GAGAAACTGGCCAAGTTTTCCGAGATTGA) was designed by screening from an siRNA expression vector library (GenoFunction, Inc., Tsukuba, Japan; patent no. 3642573 in Japan). The mitocalcin siRNA expression vector was transfected into 2Y-3t cells cultured in 10% FCS medium as described above. The transfected cells were then cultured in SF + 4F medium for neuronal differentiation.

The plasmid for the expression of mitocalcin-hemagglutinin (HA) (MCA-full) was performed as follows. The mitocalcin gene was amplified by PCR using upstream primer 5′-CTCGGATCCATGTCCAGCGAGGAGCTG-3′ containing a BamHI site and downstream primer 5′-CCGAGCTCGAGCTAAGCGTAATCTGGAACATCGTATGGGTACGCGCTGAAAGCTGCCTTGA-3′ containing an XhoI site and encoding the HA epitope. A construct of mitocalcin (55–240)-HA (MCA-ΔN) was generated by PCR amplification using upstream primer 5′-CTCGGATCCATGTCCGAACTGAACCTCAAGCTG-3′ containing a BamHI site and downstream primer 5′-CCGAGCTCGAGCTAAGCGTAATCTGGAACATCGTATGGGTACGCGCTGAAAGCTGCCTTGA-3′ containing an XhoI site and encoding the HA epitope. The products were inserted between the BamHI and XhoI sites of pcDNA3.1/Hyg. The constructs were transfected into 2Y-3t cells as described above and the cells were incubated with anti-HA epitope (Sigma Chemical Company) and anti-COX I antibodies.

Analyses of neurite length and cell death in 2Y-3t cells

The length of the longest process of 2Y-3t cells was measured with LSM Image Browser Version 3.1 (Carl Zeiss). Dying cells were also detected with 0.4% trypan blue (Sigma Chemical Company) staining. All values were presented as mean ± SEM from three independent experiments. The significance of differences between means was analysed by two-tailed Student's t-test.

Apoptosis was analysed using an annexin V-FITC apoptosis detection kit (Pharmingen, San Diego, CA, USA). In brief, 1 × 106 cells were incubated with FITC-conjugated annexin V. Propidium iodide solution was then added to the cell suspension and stained cells were analysed by a fluorescence-activated cell sorter. For DNA ladder assay, DNA was extracted from cultured cells. The DNA samples were run on 2% agarose gels and stained with ethidium bromide.

Results

Identification of mitocalcin with two EF-hand domains and a coiled-coil domain

In a previous study, we isolated genes differentially expressed in differentiated 2Y-3t cells (Tominaga and Tomooka 2002) and one of the genes encoded a novel Ca2+-binding protein named mitocalcin. The primary structure is shown in Fig. 1(a). Mitocalcin had two central EF-hand domains. We have already confirmed that the protein interacts with Ca2+ by Ca2+-binding assay (Tominaga and Tomooka 2002). In addition, the protein had a coiled-coil domain, which is known to be important in protein–protein interaction (Cohen and Parry 1990). A search for mitocalcin-related sequences showed the presence of a highly homologous gene expressed in human tissues (Fig. 1b).

Expression of mitocalcin in tissues of adult mice and developing brains

Expression of mitocalcin was examined in tissues of adult mice and developing brains with western blot analysis (Fig. 2). Mitocalcin was expressed in cerebrum, cerebellum, kidney, testis and ovary, albeit at variable levels (Fig. 2a). The protein was detected as a single band at 27 kDa. An additional band with a lower molecular weight was recognized in samples prepared from kidney, testis and ovary. In data with overexposure, additional bands were detected at a lower position than a mitocalcin band in all tissues examined and we have recently found a homolog of mitocalcin (GenBank/EMBL/DDBL Accession no. NM_025994, unpublished observations). However, the position of the homolog band was different from those of additional bands detected with anti-mitocalcin antibody. Therefore, the additional bands recognized in the samples from kidney, testis and ovary were probably caused by degradation of mitocalcin.

Figure 2.

Western bolt analysis of mitocalcin expression in mouse tissues. (a) Expression of mitocalcin in tissues of adult mouse; lanes from left to right corresponding to cerebrum, cerebellum, heart, liver, spleen, kidney, skeletal muscle, testis and ovary. (b) Expression of mitocalcin in developing brain; lanes from left to right corresponding to brain at embryonic day (E)15, post-natal day (P)0, 5, 7, 10, 15 and 20 and adult. An immunoreactive band of mitocalcin was detected at 27 kDa (arrowhead in a or b). After western blotting, the membranes were stained with ponceau solution (right panels in a and b). (c) Mitocalcin was undetected with pre-immune serum.

In developing brains, the expression was undetectable at embryonic day 15 and increased during post-natal to adult development (Fig. 2b). No immunoreactive bands of mitocalcin were detectable with rabbit pre-immune serum (Fig. 2c).

Expression of mitocalcin in 2Y-3t cells

2Y-3t is a clonal neural progenitor cell line derived from a cerebellum of an adult p53-deficient mouse (Tominaga et al. 2005). The cells cultured in 10% FCS medium take an epithelial-like morphology (Fig. 3a) and they proliferate and express nestin and Sox2, markers of neural progenitor/stem cells. When the cells are cultured in SF + 4F medium, they differentiate into mainly neurons (Fig. 3b). Expression of mitocalcin was detected in differentiated 2Y-3t cells but it was undetectable in undifferentiated 2Y-3t cells in western blot analysis (Fig. 3c). The result confirmed the previous result of mRNA expression in northern blot analysis (Tominaga and Tomooka 2002).

Figure 3.

Morphology of 2Y-3t cells and western blot analysis of mitocalcin expression in 2Y-3t cells. (a) 2Y-3t cells cultured in 10% fetal calf serum (FCS) medium took an epithelial-like morphology. (b) 2Y-3t cells cultured in serum-free plus 4 factors (SF + 4F) medium underwent striking changes in size and shape and extended long cytoplasmic processes. Scale bars, 100 µm. (c) Expression of mitocalcin in 2Y-3t cells cultured in 10% FCS (lane 1), in SF + 4F medium for 1 day (lane 2), 3 days (lane 3), 5 days (lane 4) and 10 days (lane 5). An arrowhead indicates the single band of mitocalcin.

Specific expression of mitocalcin in neurons

Expression of mitocalcin was examined immunocytochemically in differentiated 2Y-3t or FBD-102b (an oligodendrocyte cell line) and FBD-104 (an astrocyte cell line) cells established from fetal brains of p53-deficient mice (Fig. 4). Differentiated 2Y-3t cells expressing mitocalcin were positive for β-tubulin III, a neuronal marker (Fig. 4a). The expression was also observed in neurons prepared from forebrain or dorsal root ganglion in primary culture (data not shown). No expressions of mitocalcin were detected in FBD-102b (Fig. 4b) or FBD-104 (Fig. 4c) cells.

Figure 4.

Expression of mitocalcin in neural cell lines. (a) Differentiated 2Y-3t cells were stained with anti-mitocalcin (green) and anti-β-tubulin III (red, a neuronal marker) antibodies. The cell expressing mitocalcin was positive for β-tubulin III. (b) FBD-102b cells were stained with anti-mitocalcin (green) and anti-myelin basic protein (MBP) (red, an oligodendrocyte marker) antibodies. Mitocalcin was undetected in MBP-positive cells. (c) FBD-104 cells were stained with anti-mitocalcin (green) and anti-glial fibrillary acidic protein (GFAP) (red, an astrocyte marker) antibodies. GFAP-positive cells were negative for mitocalcin. Scale bars, 50 µm in (a) and (b); 100 µm in (c).

Subcellular localization of mitocalcin in neurons

Localization of mitocalcin in differentiated 2Y-3t cells was examined immunocytochemically with anti-mitocalcin antibody (Fig. 5). In differentiated 2Y-3t cells, immunoreactive signals of mitocalcin were detected in the perinuclear cytoplasm, neurites and growth cones (Fig. 5a). Observation with differential interference contrast microscopy demonstrated that immunoreactive signals of the protein were colocalized with membrane organelle-like structures in differentiated 2Y-3t cells (Figs 5b and c).

Figure 5.

Localization of mitocalcin in a differentiated 2Y-3t cell. (a) A differentiated 2Y-3t cell was stained with rhodamine-phalloidin (red) and anti-mitocalcin (green) antibodies. Mitocalcin was localized in the perinuclear cytoplasm, neurites and growth cones (arrowheads). (b) Immunofluorescence staining with anti-mitocalcin antibody in a differentiated 2Y-3t cell. (c) Differential interference contrast microscopic analysis in the differentiated 2Y-3t cell shown in (b). Immunoreactive signals of mitocalcin were well colocalized with membrane organelle-like structures (arrowheads in b and c). Scale bars, 50 µm.

Double-labeling studies were carried out with anti-mitocalcin antibody and antibodies against GM130 as a Golgi marker, GRP94 as an endoplasmic reticulum marker or COX I as a mitochondrion marker (Fig. 6). In differentiated 2Y-3t cells, none of the signals for Golgi and endoplasmic reticulum were colocalized with mitocalcin signals (Figs 6a and b). Mitocalcin signals were confined to mitochondria in differentiated 2Y-3t cells (Fig. 6c).

Figure 6.

Subcellular localization of mitocalcin in differentiated 2Y-3t cells. Double-labeling with anti-mitocalcin (green) and anti-GM130 (red) antibodies showed that the Golgi apparatus was negative for mitocalcin in differentiated 2Y-3t cells (a). Double-labeling with anti-mitocalcin (green) and anti-GRP94 (red) antibodies showed that the staining pattern of endoplasmic reticulum was distinguished from the pattern of mitocalcin staining in differentiated 2Y-3t cells (b). Double-labeling with anti-mitocalcin (green) and anti-cytochrome oxidase subunit I (COX I) (red) antibodies showed that the staining pattern of mitochondria was overlapped by the pattern of mitocalcin staining in a differentiated 2Y-3t cell (c). Scale bars, 20 µm.

To test whether mitocalcin is copurified with mitochondria, western blot analysis was conducted on subcellular fractions prepared from differentiated 2Y-3t cells (Fig. 7). To assess the functional or morphological integrity, western blot analyses were conducted with anti-COX I, anti-cytochrome c and anti-VDAC antibodies. Immunoreactive bands for each antibody were detected only in the mitochondrial fraction (upper panel in Fig. 7 and data not shown). In addition, electron microscopic analyses showed that all compartments of the mitochondria were intact (data not shown). Consistent with the immunofluorescence localization, an immunoreactive band against mitocalcin was enriched in the mitochondrial fraction (lower panel in Fig. 7).

Figure 7.

Co-purification of mitocalcin with mitochondria. Western blot analysis was conducted on subcellular fractions prepared form differentiated 2Y-3t cells. An immunoreactive band for cytochrome oxidase subunit I (COX I) was detected only in the mitochondrial fraction (arrowhead in upper panel). An immunoreactive band for mitocalcin was enriched in the mitochondrial fraction (arrowhead in lower panel). T, total cell protein; C, cytosol fraction; M, mitochondrial fraction.

Distribution of mitocalcin in mouse cerebellum

Distribution of mitocalcin was immunohistochemically examined in adult mouse cerebella (Fig. 8). In the cerebellar cortex, immunoreactive signals of mitocalcin were observed in the cytoplasm of Purkinje cells, granule cells and interneurons in the molecular layer, while the level of COX I signals was low or undetectable in the cytoplasm of those neurons (Fig. 8a). In glomeruli of the internal granular layer and in the molecular layer, signals of mitocalcin were detected and they were overlapped by the distribution of COX I signals (Fig. 8a), although mitocalcin signals were not observed in all mitochondria (Fig. 8b) at higher magnification of glomeruli in the internal granular layer.

Figure 8.

Distribution of mitocalcin in mouse cerebellar cortex. (a) Double-labeling for mitocalcin (red) and cytochrome oxidase subunit I (COX I) (green) was performed in adult cerebellum. Immunoreactivity for mitocalcin was detected in the Purkinje cell layer (PL), internal granular layer (IGL) and molecular layer (ML) of cerebellar cortex. Signals were detected in glomeruli of the IGL and in the ML and the staining pattern was well overlapped by the staining pattern of COX I (b). At higher magnification of cerebellar glomeruli, the staining pattern of COX I corresponded predominantly to that of mitocalcin (arrows). Signals of mitocalcin were detected in cytoplasm of granule cells (arrowheads). Scale bars, 20 µm in (a); 10 µm in (b).

Mitocalcin is associated with mitochondrial inner membrane

To localize with higher resolution, immunogold labeling of mitocalcin was conducted on ultrathin cryosections of adult cerebella (Fig. 9). Figure 9(a) shows mitochondria labeled with anti-mitocalcin antibody in glomeruli of the internal granular layer. Gold particles localizing mitocalcin were observed within mitochondria. Many gold particles were associated with the mitochondrial inner membrane (arrowheads in Fig. 9b). Some gold particles were also observed near to the inner membrane (arrows in Fig. 9b). The results are summarized in Fig. 9(c).

Figure 9.

Immunogold staining of mitocalcin in ultrathin cryosections of adult cerebellum. (a) Immunogold labeling for mitocalcin was detected within mitochondria. (b) A high magnification view of (a). Many gold particles for mitocalcin were localized at the inner membrane (arrowheads) or near to the inner membrane (arrows). Scale bars, 200 nm. (c) Distribution of gold particles in mitochondrial compartment. The values are shown as mean ± SEM. OM, outer membrane; IMS, intermembrane space; IM, inner membrane; ND, not determined.

Overexpression of mitocalcin in 2Y-3t cells

To examine the function of mitocalcin, 2Y-3t cells were transfected with pcDNA3.1-sense mitocalcin vectors and S5 and S43 sublines were established. Western blot analysis showed that mitocalcin was expressed in both S5 and S43 cells cultured in 10% FCS medium and the expression was undetected in controls (2Y-3t cells or V26 and V38 cells transfected with pcDNA3.1 vector) (Fig. 10a). Western blot analysis was performed on subcellular fractions prepared from S43 cells cultured in 10% FCS medium. An immunoreactive band of mitocalcin was enriched in the mitochondrial fraction compared with that of mitocalcin in the cytosol fraction (Fig. 10b).

Figure 10.

Overexpression of mitocalcin in 2Y-3t cells. (a) Western blot analysis showed that mitocalcin was expressed in S5 and S43 cells cultured in 10% fetal calf serum (FCS) medium. Mitocalcin was undetected in control cells (parental 2Y-3t, V26 and V38 cells). (b) Co-purification of mitocalcin with mitochondria fractionated from V26 or S43 cells cultured in 10% FCS medium. Cytochrome oxidase subunit I (COX I) was detected in the mitochondrial fraction (lane M) of both V26 and S43 cells. Mitocalcin was enriched in lane M of S43 cells but the protein was undetected in subcellular fractions from V26 cells. (c) Phase-contrast photomicrographs of V26 (upper panel) and S43 (lower panel) cells. The longest process of an S43 cell is indicated between two arrowheads in the lower panel. Scale bars, 50 µm. (d) The length of the longest processes in mitocalcin-overexpressing cells was approximately 2.7-fold longer than that of control cells. The values showed as mean ± SEM (*p < 0.01).

Overexpression of mitocalcin induced morphological changes in 2Y-3t cells. 2Y-3t cells cultured in 10% FCS medium had mainly flat cytoplasmic processes. The processes became longer and slender like neurites in S5 and S43 cells cultured in 10% FCS medium (Fig. 10c). The length of the longest process in the overexpressing cells was approximately 2.7-fold longer than that of control cells (Fig. 10d).

Inhibition of mitocalcin expression in 2Y-3t cells

In an effort to clarify the function, inhibition of mitocalcin expression was attempted by establishing sublines A16 and A48 expressing antisense mitocalcin mRNA (Fig. 11). Western blot analysis showed that mitocalcin expression was reduced by approximately 75% in both A16 and A48 cells cultured in SF + 4F medium for 3 days (Fig. 11a). Inhibition of mitocalcin expression caused a reduction of neurite extension (Fig. 11b).

Figure 11.

Inhibition of mitocalcin expression in 2Y-3t cells. (a) Expression of mitocalcin in control cells (parent 2Y-3t, V26 and V38 cells) and in sublines expressing antisense mitocalcin mRNA (A16 and A48 cells) cultured in SF + 4F medium for 3 days. The expression level in A16 and A48 cells was reduced. (b) Phase-contrast photomicrographs of V26 (left panel) and A48 (right panel) cells cultured in SF + 4F medium for 3 days. Expression of antisense mitocalcin mRNA in A48 cells decreased the number of cells extending neurites compared with V26 cells. Scale bar, 100 µm. (c) Trypan blue staining of control cells and cells expressing antisense mRNA cultured in SF + 4F medium for 5 days. The percentage of stained cells is shown as mean ± SEM (*p < 0.01). (d) DNA ladder formation in antisense-expressing cells cultured in SF + 4F medium for 5 days. DNA was extracted and run on 2% agarose gels and stained with ethidium bromide. DNA ladders were undetected in both control (lane 1, V26; lane 2, V38) and antisense-expressing (lane 3, A16; lane 4, A48) cells. Lanes: M, DNA marker; 5, sample from K2 cells treated with tamoxifen (Iida et al. 1998) as a positive control.

A trypan blue exclusion test was conducted to determine whether the knockdown affects cell viability. Cells were cultured in SF + 4F medium for 5 days and stained with trypan blue solution. The result showed that numbers of stained cells in the sublines were larger than those of controls (Fig. 11c).

2Y-3t cells were transfected with a mitocalcin siRNA expression vector to exclude the risk of non-specific knockdown of accessory genes. The same results were obtained as in the full-length antisense experiments (Tominaga et al., unpublished observations).

To determine whether the cell death was caused by apoptosis, DNA laddering was examined in antisense-expressing cells. DNA laddering was not observed in A16 and A48 cells cultured in SF + 4F medium for 5 days (Fig. 11d). An annexin V assay was also conducted in antisense-expressing cells. The number of apoptotic cells (annexin V-positive/propidium iodide-negative) in antisense-expressing cells (A16 and A48) was not significantly different from the number in control cells (V26 and V38) (data not shown).

Discussion

In this study, we found that the mitocalcin gene encoded a novel mitochondrial Ca2+-binding protein. Although our previous study showed that the recombinant mitocalcin interacts with Ca2+in vitro (Tominaga and Tomooka 2002), this property of mitocalcin was not confirmed in mitochondria. We expected that mitochondria in cells overexpressing mitocalcin would have a significantly higher capacity of calcium retention if the protein in mitochondria interacts with Ca2+. To assess the interaction with Ca2+, mitocalcin-overexpressing cells were stained with magfura-2, which is a fluorescent dye with low affinity for Ca2+, and incubated with the protonophore carbonylcyanide-p-(trifluoromethoxy)-phenylhydrazone to release mitochondrial Ca2+ into the cytoplasm (Brocard et al. 2001). However, a significant change in the fluorescence intensity of cytosolic free calcium was not detected between mitocalcin-overexpressing and control (only vector-expressing) cells (data not shown). This result suggests the possibility that mitocalcin binds with Ca2+ only when it is activated in mitochondria or the interaction has a low affinity with Ca2+ in mitochondria. The issue should be re-examined with other methods.

Analyses of mitocalcin expression showed that neurons expressed mitocalcin and the expression in developing brain was increased during post-natal to adult development. These data suggest that mitocalcin is involved in neuronal maturation rather than neurogenesis. In addition, functional changes from a low to high metabolism are known in mitochondria in brain during early post-natal development (Thurston and McDougal 1969). The developmental expression of mitocalcin implies involvement of the protein in mitochondrial energy metabolism.

Mitocalcin was expressed specifically in differentiated neurons and it was not expressed in undifferentiated neural cells or glial cells in cell lines. The specific expression was confirmed by immunohistochemistry of adult cerebellar cortices. Anti-mitocalcin staining was seen in all classes of neurons, including both inhibitory (e.g. Purkinje cells) and excitatory (e.g. granule cells) types. Immunohistochemistry of other regions of brain also demonstrated that mitocalcin expression was detected in neurons but it was undetected in glial cells (data not shown).

Immunogold labeling showed that mitocalcin was associated with the mitochondrial inner membrane. When immunostaining for mitocalcin was conducted in fixed cells or tissue sections untreated with TritonX-100, the signals were very weak or undetected (data not shown), supporting the idea that mitocalcin localizes within mitochondria but is not associated with their outer membrane. However, the protein was not seen in all mitochondria of cerebellar neurons. This observation may imply that mitochondria in brain are functionally heterogeneous, as demonstrated by biochemical studies of brain mitochondria in terms of enzyme composition (Blokhuis and Veldstra 1970; Wilkin et al. 1979).

Typical mitochondrial targeting signals are encoded in the N-terminal pre-sequences, which have the potential to form positively charged amphiphilic helices (Roise and Schatz 1988). Such targeting signals were not predicted in mitocalcin protein with PSORT or MitoProt programs (data not shown) (Emanuelsson and von Heijne 2001). A construct of mitocalcin (55–240)-HA (MCA-ΔN) was generated by PCR amplification. 2Y-3t cells cultured in 10% FCS medium were transfected with the construct and stained with anti-HA tag and anti-COX I antibodies. MCA-ΔN was localized in mitochondria (data not shown), suggesting that the mitochondrial target signal of mitocalcin may be in other regions except for the N-terminal of the protein.

Mitocalcin was localized in growth cones of differentiated 2Y-3t cells and in cerebellar glomeruli which contain many synapses (Palay and Chan-Palay 1974). Such outgrowths of cytoplasm dynamically change morphology and function under Ca2+ regulation (Kater and Mills 1991; Mattson 1992; De Camilli and Takei 1996) and much energy is consumed in neurite extension (Bridgman et al. 1994; Desai et al. 1997) and synaptic functions (Bernstein et al. 1998; Fischer et al. 1998). Mitochondria are involved in Ca2+ regulation and energy supply in the structures (Tang and Zucker 1997; Bernstein and Bamburg 2003; Levy et al. 2003) and Ca2+ is believed to be one of the regulatory factors of oxidative phosphorylation (McCormack et al. 1990; McCormack and Denton 1993). Our data showed that overexpression of mitocalcin resulted in neurite extension in 2Y-3t cells, while suppression of the expression caused neurite reduction and then cell death. Activation of the apoptotic pathway in antisense-expressing cells was undetected in assays of DNA ladder formation and annexin V staining. Therefore, the reduced viability of antisense-expressing cells may be a consequence of the reduction of neurite extension, suggesting that mitocalcin might play roles in maintaining neuronal function through mitochondrial energy metabolism. This conclusion is supported by recent reports. Neuronal death is caused by the dysfunction of energy metabolism (Atorino et al. 2003; Lindholm et al. 2004) and mitochondrial dysfunction is a primary or secondary factor in many diseases, including neurodegenerative diseases (Wallace 1999; Schapira 2000; Schon 2000).

Acknowledgements

Financial assistance was provided by Specific Coordination Funds (SPSBS) of the Japanese Ministry of Education, Sports, Science and Technology and by MEXT. HAITEKU (2003). We thank all members of the Tomooka and Sakai laboratories for helpful comments and discussion.

Ancillary