Functional and immunocytochemical evidence for the expression and localization of the secretory pathway Ca2+-ATPase isoform 1 (SPCA1) in cerebellum relative to other Ca2+ pumps

Authors

  • M. Rosario Sepúlveda,

    1. Departamento de Bioquímica y Biología Molecular y Genética, Facultad de Ciencias, Universidad de Extremadura, Badajoz, Spain
    2. Laboratorium voor Fysiologie, Katholieke Universiteit Leuven, Campus Gasthuisberg, Leuven, Belgium
    Search for more papers by this author
  • María Berrocal,

    1. Departamento de Bioquímica y Biología Molecular y Genética, Facultad de Ciencias, Universidad de Extremadura, Badajoz, Spain
    Search for more papers by this author
  • Daniel Marcos,

    1. Departamento de Bioquímica y Biología Molecular y Genética, Facultad de Ciencias, Universidad de Extremadura, Badajoz, Spain
    Search for more papers by this author
  • Frank Wuytack,

    1. Laboratorium voor Fysiologie, Katholieke Universiteit Leuven, Campus Gasthuisberg, Leuven, Belgium
    Search for more papers by this author
  • Ana M. Mata

    1. Departamento de Bioquímica y Biología Molecular y Genética, Facultad de Ciencias, Universidad de Extremadura, Badajoz, Spain
    Search for more papers by this author

Address correspondence and reprint requests to Ana M. Mata, Departamento de Bioquímica y Biología Molecular y Genética, Facultad de Ciencias, Universidad de Extremadura, 06071 Badajoz, Spain.
E-mail: anam@unex.es

Abstract

Membrane fractions of pig cerebellum show Ca2+-ATPase activdity and Ca2+ transport due to the presence of the secretory pathway Ca2+-ATPase (SPCA). The SPCA1 isoform shows a wide distribution in the neurons of pig cerebellum, where it is found in the Golgi complex of the soma of Purkinje, stellate, basket and granule cells, and also in more distal components of the secretory pathway associated with a synaptic localization such as in cerebellar glomeruli. The SPCA1 may be involved in loading the Golgi complex and the secretory vesicles of these specific neuronal cell types with Ca2+ and also Mn2+. This study of the cellular and subcellular localization of SPCA1 pumps relative to the sarco(endo) plasmic reticulum Ca2+-ATPase and plasma membrane Ca2+-ATPase pumps hints to a possible specific role of SPCA1 in controlling the luminal secretory pathway Ca2+ (or Mn2+) levels as well as the local cytosolic Ca2+ levels. In addition, it helps to specify the zones that are most vulnerable to Ca2+ and/or Mn2+ dyshomeostasis, a condition that is held responsible of an increasing number of neurological disorders.

Abbreviations used
DAPI

4′,6-diamidino-2-phenylindole

GFAP

glial fibrillary acidic protein

PBS

Phosphate buffered saline

PMCA

plasma membrane Ca2+-ATPase

PVDF

polyvinylidene difluoride

SERCA

sarco(endo)plasmic reticulum Ca2+-ATPase

SPCA

secretory pathway Ca2+-ATPase

TBS

Tris buffered saline

The secretory pathway Ca2+-ATPases (SPCA) are intrinsic membrane proteins localized in the membranes of the Golgi complex and the secretory pathway of mammalian cells. They represent a third type of P-type Ca2+-transport ATPases, besides the Ca2+-ATPases of sarco(endo)plasmic reticulum (SERCA) and plasma membrane (PMCA). The SPCAs share with SERCA and PMCA proteins the ability to transport Ca2+. In addition the SPCAs can also transport Mn2+ with an affinity similar to that for Ca2+ (Van Baelen et al. 2001; Ton et al. 2002). Two genes (ATP2C1 and ATP2C2) encoding respectively the SPCA1 (Hu et al. 2000; Sudbrak et al. 2000) and SPCA2 (Xiang et al. 2005) isoforms have been identified in higher vertebrates. The SPCA1 isoform is expressed in many cell types and it represents a house-keeping isoform. Interestingly, human keratinocytes, more than other cells, are extremely sensitive to a disturbance of SPCA1 function. Indeed, mutations in a single ATP2C1 allele result in a skin disorder known as Hailey-Hailey disease (Hu et al. 2000; Sudbrak et al. 2000). However, SPCA2 seems to have a more restricted tissue distribution than SPCA1 and its biological function is not known yet (Vanoevelen et al. 2005; Xiang et al. 2005).

In the nervous system, Ca2+ ions play a key role in neural plasticity, synaptic transmission, neuronal ageing, or apoptosis (Squier and Bigelow 2000). Thus, alterations in Ca2+ homeostasis are one of the more serious causes of neurological disorders (Mattson et al. 2000). The first evidence for the presence of SPCA in the nervous system was recently reported by Wootton et al. (2004), Xiang et al. (2005), and Murin et al. (2006) but a detailed assignment of this protein to membrane fractions, its physiological localization in neural tissue, and knowledge on the particular functions of SPCA in the nervous system are currently lacking. In this work, we show for the first time the presence of the SPCA1 isoform in pig cerebellar tissue and its physiological distribution in specific cellular types. Besides, we try to assess the functional contribution of SPCA to the total ATP hydrolysis and Ca2+ transport catalyzed by the three types of Ca2+ pumps in pig cerebellar membranes.

Materials and methods

Materials

Thapsigargin and ammonium vanadate were obtained from Sigma (Madrid, Spain). 45Ca2+ was from American Radiolabeled Chemicals, Inc. (St Louis, MO, USA), and the scintillation fluid Filtron-X was from National Diagnostics (Atlanta, GE, USA). The polyclonal rabbit anti-SPCA1 (hSPCA1cytl) was prepared against the large cytosolic loop of human SPCA1 as described earlier (Van Baelen et al. 2003). The SERCA2-specific monoclonal antibody IID8 and the PMCA-specific monoclonal antibody 5F10 were purchased from Affinity Bioreagents (Golden, CO, USA), the anti-γ adaptin, the anti-glial fibrillary acidic protein (GFAP) and the anti-β tubulin monoclonal antibodies were from Sigma, and the anti-synaptophysin monoclonal antibody was from CRP Inc. (Denver, PA, USA). The secondary antibodies conjugated with peroxidase were purchased from BioRad (Madrid, Spain), those biotinylated and the ExtrAvidin-peroxidase were from Sigma, and the fluorescent secondary antibodies labeled with Alexa 488 or Alexa 594 were obtained from Molecular Probes (Eugene, OR, USA). All other reagents were of the highest purity available.

Preparation of subcellular fractions from pig cerebellum

Subcellular fractions were prepared from adult pig (5 months-old) cerebellum following the procedure described by Salvador and Mata (1996) for pig brain. Briefly, four fresh cerebella, obtained from the local slaughterhouse, were homogenized in 10 volumes of 10 mmol/L Hepes/KOH pH 7.4, 0.32 mol/L sucrose, 0.5 mmol/L MgSO4, 0.1 mmol/L phenylmethanesulphonyl fluoride, and 2 mmol/L 2-mercaptoethanol. After two centrifugation steps at low and high speed respectively, microsomes (enriched in intracellular membranes) were isolated from the supernatant. Synaptosomes were obtained from the pellet using a discontinuous sucrose gradient and collected at the 20–40% interface.

Ca2+-ATPase activity

The enzymatic activity was measured at 37°C by using a coupled enzyme assay, following spectrophotometrically the NADH absorption change at 340 nm. Briefly, 40 μg of pig cerebellum fractions were incubated for 4 min with a reaction mixture containing, in a final volume of 1 mL, 50 mmol/L Hepes/KOH pH 7.4, 100 mmol/L KCl, 100 μmol/L CaCl2, 2 mmol/L MgCl2, 5 mmol/L NaN3, 100 μmol/L 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (3.16 μmol/L free Ca2+), 0.22 mmol/L NADH, 0.01% saponin (to disrupt both the plasma and intracellular membranes), 0.42 mmol/L phosphoenolpyruvate, 10 IU of pyruvate kinase, 28 IU of lactate dehydrogenase (Sepúlveda et al. 2005). The reaction was started by the addition of 1 mmol/L ATP. The further addition of 100 nmol/L thapsigargin (to inhibit SERCA activity), and of 2 μmol/L vanadate (a dose used to selectively inhibit PMCA activity, freshly prepared to avoid its transformation to decameric vanadate species) and of 3 mmol/L EGTA (to measure the Mg2+-ATPase activity) were made in three different steps, and the activity was measured after each addition in order to evaluate the contribution of each ATPase to the total ATPase activity. The total Ca2+-ATPase activity (i.e. the sum of SERCA, PMCA, and SPCA) was obtained after subtraction of the Mg2+-ATPase activity from the (Ca2++Mg2+)-ATPase activity obtained immediately after the addition of ATP. The SERCA activity was calculated by subtracting the activity measured in the presence of thapsigargin (which includes PMCA, SPCA, and Mg2+-ATPase activities) from the total (Ca2++Mg2+)-ATPase activity. The SPCA activity was calculated by subtracting the Mg2+-ATPase activity from the activity in the presence of thapsigargin and vanadate. The PMCA activity was calculated by subtracting the Mg2+-ATPase activity and the SPCA activity from the activity in the presence of thapsigargin.

Ca2+ uptake

The 45Ca2+ transport was measured at 37°C by the filtration technique (Chiesi and Inesi 1979). Subcellular fractions (40 μg) were incubated in an uptake medium containing 50 mmol/L Hepes/KOH pH 7.4, 100 mmol/L KCl, 100 μmol/L CaCl2, 2 mmol/L MgCl2, 5 mmol/L NaN3, 100 μmol/L 1,2-bis(2-aminophenoxy)ethane-N,N,N’,N’-tetraacetic acid (to obtain Ca2+ free 3.16 μmol/L), 20 mmol/L KH2PO4, 0.3 μCi of 45CaCl2 (approximately 20 000 cpm/nmol) in 1 mL total volume. After 4 min incubation at 37°C, the uptake reaction was started by addition of 1 mmol/L ATP and was stopped 30 min later or at the time specified by filtration through Millipore filters (HAWP-045; Millipore, Madrid, Spain). Filters were washed three times with 3 mL each of 20 mmol/L Hepes and 1 mmol/L LaCl3. Filters were thereupon dissolved in scintillation fluid (Filtron-X, National Diagnostics), and counted. One blank without protein were filtrated to determine the level of non-specific calcium retained by the filter.

SERCA, PMCA, and SPCA all contributed to the Ca2+ uptake. SPCA-mediated 45Ca2+ uptake was measured after addition to the medium of 100 nmol/L thapsigargin (to inhibit SERCA) and of 0.002% saponin, to selectively disrupt the plasma membrane and hence the PMCA-mediated transport (Missiaen et al. 1994; Van Baelen et al. 2001). The Ca2+ transport by SERCA was calculated by subtraction of the values obtained in the presence of thapsigargin from the total value found without thapsigargin and saponin. The Ca2+ transport by PMCA was calculated by subtraction of SPCA Ca2+ transport from that in the presence of thapsigargin.

Protein determination

The protein content was evaluated by the Bradford (1976) method using bovine serum albumin as a standard.

Electrophoresis and immunoblotting

Electrophoresis was performed by the method of Laemmli (1970) in 6.5% (w/v) polyacrylamide gels. Protein transfer to a polyvinylidene difluoride (PVDF) membrane was carried out in a Trans-Blot® SD electrophoretic transfer cell semi-dry system (Bio-Rad). The PVDF membrane was blocked for 30 min in Tris buffered saline (TBS) containing 2% (w/v) non-fat dry milk (TBS-milk). Immunological reactions were performed by incubation of the membrane with the primary antibody for 2 h at 25°C, at the following dilutions in TBS containing 0.05% (v/v) Tween 20: the polyclonal anti-SPCA1 (1 : 1500), anti-γ adaptin (1 : 200), and the anti-β tubulin (1 : 1000) antibodies. Afterwards, the membrane was incubated for 1 h at 25°C with peroxidase-conjugated secondary antibody and stained with 4-methoxy-1-naphthol. The PVDF membrane was washed extensively with TBS-milk between the different incubations. In addition to the standard western blot, a special Odyssey® Infrared Imaging System (LI-COR Biosciences, Lincoln, NE, USA) has been used to detect two separate proteins in the same assay. Briefly, the membrane was blocked in Odyssey Blocking Buffer for 1 h, and then incubated with the primary antibodies anti-SPCA1 (1 : 1500) and anti-γ adaptin (1 : 200) diluted in the same Blocking Buffer, for 2 h at 25°C. After three washes with 0.1% Tween 20 in phosphate buffer saline (PBS) for 5 min, the membrane was incubated with a mixture of two Infrared (IR)-labeled secondary antibodies (goat anti-mouse IRDye™ 800 (LI-COR Biosciences) and goat anti-rabbit Alexa Fluor®680 (Molecular Probes)), diluted 1 : 5000 in the Odyssey Blocking Buffer, for 1 h at 25°C in the dark. The membrane was washed with 0.1% Tween 20 in PBS, was kept at 4°C in dark and visualized with the Odyssey scanner.

Tissue preparation for immunohistochemistry

Fresh adult pig cerebella were divided into slabs in a sagittal plane and fixed by immersion for 24 h at 4°C in 4% (w/v) paraformaldehyde in PBS solution. The pieces were rinsed in PBS with several changes for 2 days, cryoprotected in 10% (w/v) sucrose in PBS for 2 days, and then embedded in 10% (w/v) gelatine, 10% (w/v) sucrose in PBS. The blocks were frozen for 2 min in isopentane cooled to −70°C with dry ice, and stored at −80°C. Serial parasagittal 20 μm-thick sections were collected on Super-Frost Plus slides using a Leica CM1900 cryostat (Barcelona, Spain).

Immunohistochemistry

The localization of the SPCA1 protein in sections was performed according to Sternberger et al. (1970), using the anti-SPCA1 antibody (1 : 500). After several washes, the sections were incubated with biotynilated goat anti-rabbit (1 : 200) and then with ExtrAvidin-peroxidase (1 : 200). The immunodetection of the peroxidase activity was carried out using 0.03% (w/v) 3,3′-diaminobenzidine tetrahydrochloride. The sections were mounted with Eukitt for observation under the microscope.

The dual localization of proteins was done by incubating the sections with a mixture of rabbit anti-SPCA1 antibody (1 : 500) and one of the mouse monoclonal antibodies anti-γ adaptin (1 : 200), anti-GFAP (1 : 400), IID8 (1 : 500), 5F10 (1 : 500) or anti-synaptophysin (1 : 500), or with 10 μmol/L 4′,6-diamidino-2-phenylindole (DAPI). Fluorescence labeling was observed using two secondary antibodies (Alexa594 goat anti-rabbit and Alexa488 goat anti-mouse, 1 : 500). Thereupon, the sections were covered with FluorSave mounting medium and analyzed using a Nikon E600 microscope (Melville, NY, USA). Negative controls were performed for every set of experiments by omitting the primary antibodies from the procedure.

Results

Distribution of SPCA1 in subcellular fractions of pig cerebellum

The presence and expression of SPCA1 protein in microsomes and synaptosomes from pig cerebellum was shown by western-blot assays (Fig. 1a). The specific anti-SPCA1 antibody labeled a protein band around 100 kDa in both membrane preparations. The absence of cross-reactivity against SPCA2 was confirmed using microsomes isolated from COS cells over-expressing the SPCA2 isoform. These fractions also contained the γ adaptin protein, i.e. a marker for Golgi compartments and trans-Golgi network. Identical amounts of total protein were loaded in each lane, as revealed by the anti-β tubulin antibody. Both proteins, γ adaptin and SPCA, were also observed using the Odyssey system (Fig. 1b) which allows the simultaneous demonstration of the two proteins each labeled with a different fluorescent secondary antibody.

Figure 1.

 Immunoreaction of secretory pathway Ca2+-ATPase isofom 1 (SPCA1) and γ adaptin with their specific antibodies in subcellular fractions of pig cerebellum. (a) Forty micrograms of microsomes (MIC) or synaptosomes (SYN), and 5 μg de microsomes from COS cells over-expressing SPCA1 or SPCA2. The polyvinylidene difluoride membranes were incubated with the indicated antibodies and revealed with 4-methoxy-1-naphthol. The β tubulin was used as a loading control. (b) Similar immunoreactions were performed using the Odyssey assay, which allows the simultaneous visualization of both proteins by immunofluorescence (see Materials and methods section).

Functional expression of SPCA in cerebellar membrane fractions

Measurements of Ca2+-ATPase activity (Fig. 2a) and Ca2+ uptake (Fig. 2b) were performed in microsomes and synaptosomes using selective conditions, as described in the Materials and methods section, in order to demonstrate the functional presence of SPCA in cerebellum and assess its contribution to the total activity with respect to that of SERCA and PMCA. SPCA-mediated Ca2+-ATPase and 45Ca2+-transport activities were comparable in both fractions. However, the corresponding values for SERCA and PMCA were found to differ markedly. SERCA was approximately 1.6 fold higher in microsomes, whereas PMCA was about 2.1–3.5 fold higher in synaptosomes. Similar values were obtained when 2 μmol/L vanadate was included in the assay in order to inhibit the Ca2+ transport by PMCA (results not shown).

Figure 2.

 Sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA)-, secretory pathway Ca2+-ATPase (SPCA)- and plasma membrane Ca2+-ATPase (PMCA)-mediated ATPase activity and 45Ca2+ uptake in subcellular fractions. Forty micrograms of microsomes or synaptosomes were incubated at 37°C for 4 min and the reaction was started with 1 mmol/L ATP. The measurements of activity (a) and 45Ca2+ uptake (b) catalyzed by each type of Ca2+ pump were performed as indicated in the Materials and methods section. Data are mean ± SE (bars) values obtained from five different experiments, performed in triplicate, and from three different preparations.

The SPCA is the only Ca2+ pump amongst the P-type Ca2+-ATPases in this preparation that can also transport Mn2+ ions. In fact, the addition of increasing concentrations of Mn2+ resulted in a parallel decrease of the SPCA-mediated 45Ca2+ uptake (Fig. 3), but not of the SERCA-mediated nor of the PMCA-mediated components (Table 1). Besides, it is clear that higher Ca2+ concentrations require more Mn2+ to inhibit the SPCA-mediated transport of Ca2+, which was not the case for the other Ca2+ pumps. IC50 values of 0.06, 0.05, and 0.08 μmol/L Mn2+ were obtained at 0.316, 1, and 3.16 μmol/L Ca2+ free, respectively.

Figure 3.

 Mn2+ inhibition of the secretory pathway Ca2+-ATPase-mediated 45Ca2+ uptake in cerebellar synaptosomes. Forty micrograms of synaptosomes were incubated at 37°C for 4 min and the reaction was started with 1 mmol/L ATP. The 45Ca2+ uptake mediated by secretory pathway Ca2+-ATPase was assayed as indicated in the Materials and methods section in the presence of the Mn2+ concentrations indicated on the abscissa at 0.316 μmol/L (bsl00066), 1 μmol/L (bsl00001), and 3.16 μmol/L (•) free Ca2+. Data are mean ± SE (bars) values obtained from six different experiments, and from three different preparations. The 100% uptake corresponds to 21.0 ± 1.4, 18.8 ± 1.2, and 14.2 ± 0.2 nmol Ca2+/mg of protein in the presence of 0.316 μmol/L, 1 μmol/L, and 3.16 μmol/L free Ca2+, respectively.

Table 1.   Mn2+ dependence of the sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA)- and plasma membrane Ca2+-ATPase (PMCA)-mediated 45Ca2+ uptake in cerebellar synaptosomes at different Ca2+ concentrations
 SERCAPMCA
[Mn2+] (μmol/L)0101
[Ca2+] free    
  1. Ca2+ uptake of SERCA and PMCA were assayed in synaptosomes (40 μg) as indicated in the Materials and methods section in the presence of different manganese concentrations at 0.316 μmol/L, 1 μmol/L, and 3.16 μmol/L Ca2+ free. Data are mean ± SE (bars) values obtained from four different experiments, and from two different preparations.

0.316 μmol/L11.0 ± 0.211.0 ± 1.7 31.1 ± 0.03 31.6 ± 0.5
1.000 μmol/L11.0 ± 2.111.1 ± 0.4 31.2 ± 2.30 30.8 ± 2.6
3.160 μmol/L15.0 ± 0.315.1 ± 0.933.6 ± 3.0033.7 ± 2.4

Localization of SPCA1 in para-sagittal sections of pig cerebellum relative to the distribution of other Ca2+ pumps

The SPCA1 protein was localized in cryostat sections of pig cerebellum by immunohistochemical assays, using 3,3′-diaminobenzidine tetrahydrochloride (Fig. 4). The staining with an anti-SPCA1 antibody was detected in the three layers of the cerebellar cortex (Fig. 4a), in somas of Purkinje cells and of interneurons located at all levels of the molecular layer and in overall granular layer (Fig. 4b). A detailed analysis using higher magnifications of the upper molecular layer shows SPCA1-labeling in stellate cells (Fig. 4c). The staining was found in definite spots in the cytoplasm of these cells (Fig. 4d), reminiscent of a Golgi complex distribution. In the deeper half of molecular layer (Fig. 4e), a similar labeling of definite spots in the cytoplasm of the body of basket and Purkinje cells was observed. The dendritic arborization, the axon, and the nucleus of these cells were always negative. The SPCA1 labeling was also found throughout the granular layer (Fig. 4f).

Figure 4.

 Localization of secretory pathway Ca2+-ATPase isofom 1 (SPCA1) in para-sagittal sections of cerebellar cortex. The immunoreaction of SPCA1 was detected in the three layers of the cortex (a, b), in somas of Purkinje cells [(b), large arrow], in somas of interneurons located at all levels of molecular layer [(b), short arrows and arrowheads], and in the granular layer. A higher magnification of the upper molecular layer shows the SPCA1 labeling in stellate cells [(c), arrows]. The staining [high magnification in (d)] was found in definite dots in the cytoplasm of the body of these cells, but not in dendrites and nucleus. A high magnification of the deeper half of molecular layer (e) showed similar labeling in basket cells (arrowheads) and in Purkinje cell soma (arrows). The SPCA1 labeling was also found throughout the granular layer (f). ml, molecular layer; pcl, Purkinje cell layer; gl, granular layer; wm, white matter. Scale bars: a, 140 μm; b, 40 μm; c, 20 μm; d, 4,5 μm; e, 9 μm; f, 25 μm.

To analyze in more detail the specific localization of SPCA1, double immunofluorescence assays were performed. In the soma of Purkinje cells and basket cells (Fig. 5), the distribution of the Golgi marker adaptin (Fig. 5a) was in general more homogeneous than the spot-like SPCA1 staining, but it was still restricted to the cytoplasm of the bodies. The γ adaptin antibody reacts with a protein of the trans-Golgi network, probably found in all vesicles generated in the Golgi apparatus and distributed over the entire somatic cytoplasm of these neurons. Immunolabeling of SPCA1 or γ adaptin was not observed in any other region of these neurons, such as the dendritic arborization. Similar labeling to anti-γ adaptin was found with another Golgi marker, the anti-58 K Golgi protein (results not shown). Co-staining of sections with the anti-SPCA1 antibody, the glial marker anti-GFAP and DAPI (Fig. 5b) did not show a GFAP expression in SPCA1 positive cells.

Figure 5.

 Double immunolocalization of secretory pathway Ca2+-ATPase isofom 1 (SPCA1), γ adaptin, glial fibrillary acidic protein (GFAP), sarco(endo)plasmic reticulum Ca2+-ATPase 2 (SERCA2) and plasma membrane Ca2+-ATPase (PMCA) Ca2+ pumps, and synaptophysin in Purkinje and molecular layers. The anti-SPCA1 antibody [red in (a–e)] labeled definite dots located in the cytoplasm of the bodies of Purkinje cells (large arrows) and of interneurons such as basket cells (arrowheads). Double immunostaining of SPCA1 [red in (a)] and the Golgi marker γ adaptin [green in (a)] showed co-localization of these proteins (merge) in the soma of both, Purkinje cells (large arrow) and interneurons (arrowhead). The co-staining of similar sections with anti-SPCA1 [red in (b)], anti-GFAP [green in (b)] and 4′,6-diamidino-2-phenylindole (DAPI) [blue in (b)] showed that SPCA1 positive-cells (large arrow and arrowhead) did not correspond with glial cells. The SERCA2-specific IID8 antibody [green in (c)] homogeneously reacted in the cytoplasm of the soma and dendritic trees (short arrow) of Purkinje cells, thus showing co-localization with SPCA1 only in the soma (merge). The PMCA specific 5F10 antibody [green in (d)] shows a different localization from SPCA1 in Purkinje cells. PMCA is present in the periphery of the soma and dendritic spines of the whole Purkinje arborization. The synapse marker synaptophysin was found in the three layers of the cerebellar cortex [green in (e)], but did co-localize with SPCA1 neither in Purkinje cells nor in basket cells. Scale bar: 27 μm.

Of particular interest is the comparison of the distribution of SPCA1 in relation to the other Ca2+pumps which are also found in pig cerebellum, each with its own specific distribution (Sepúlveda et al. 2004). Thus, the SERCA2 protein stained with the SERCA2-specific IID8 antibody (Fig. 5c) was found homogeneously spread over the soma and dendritic trees of Purkinje cells. SERCA2 and SPCA1 were therefore only co-localized in the soma, as was also the case for γ adaptin and SPCA1. Although the immunoreaction with the SERCA2-specific IID8 in stellate and basket cells is very weak or absent, Y/1 F4, another anti-SERCA2-specific antibody, labels the soma of stellate and basket cells with a pattern similar to that of IID8 in Purkinje cells (Sepúlveda et al. 2004). The PMCA pump was labeled with the 5F10 antibody (Fig. 5d) at the periphery of the soma and in the dendritic spines of the entire Purkinje arborisation, and hence there was little co-localization with SPCA1 in these cells. The synapse marker anti-synaptophysin (Fig. 5e) reacts in the three layers of the cerebellar cortex, but did co-localize with SPCA1 neither in Purkinje nor in basket cells. In the granular layer (Fig. 6), each specific antibody labeled the perikaryon of granule cells (arrowheads) and cerebellar glomeruli (asterisks), proving the presence of SPCA1, the trans-Golgi marker γ adaptin, SERCA2 and PMCA proteins respectively in these cells. The SPCA1 colocalized with γ adaptin in the perikaryon of granule cells and in some areas of cerebellar glomeruli (merge in Fig. 6a). SPCA1 and the other calcium pumps SERCA2 and PMCA did not colocalize in all population of granule cells (merge in Fig. 6c and d) and were co-expressed in cerebellar glomeruli (merge in Fig. 6c and d). However, the SPCA1 expression was restricted to some areas of cerebellar glomeruli and in the vicinity of these structures, while the other proteins were homogeneously distributed (Fig. 6a and d). The co-expression of SPCA1 and the synapse marker synaptophysin in the cerebelar glomeruli (Fig. 6e) shows the presence of the SPCA1 in these rich-synapses structures. The DAPI-positive nucleus showed that not all cells were SPCA1 positive (Fig. 6b).

Figure 6.

 Double immunolocalization of secretory pathway Ca2+-ATPase isofom 1 (SPCA1), γ adaptin, sarco(endo)plasmic reticulum Ca2+-ATPase 2, plasma membrane Ca2+-ATPase and synaptophysin in the granular layer. The SPCA1 labeling [red in (a–d)] was found overall the granular layer. The soma of granule cells [arrowheads in (a,c,d)] was immunopositive for anti-SPCA1, anti-γ adaptin [green in (a)], IID8 [green in (c)], and 5F10 [green in (d)] antibodies. The SPCA1 colocalized with γ adaptin in the perikaryon of granule cells and in some areas of cerebellar glomeruli (merge in Fig. 6a). SPCA1 did not colocalize with sarco(endo)plasmic reticulum Ca2+-ATPase 2 and plasma membrane Ca2+-ATPase in all population of granule cells (arrowheads, merge in Fig. 6c and d), but the three ATPases were co-expressed in cerebellar glomeruli (asterisks, merge in Fig. 6c and d). The staining of SPCA1 was concentrated in some areas within and around of the cerebellar glomeruli while the anti-γ adaptin (a), IID8 (c), 5F10 (d), and anti-synaptophysin (e) antibodies labeling were homogeneous (merge). Sections stained with 4′,6-diamidino-2-phenylindole (DAPI) [blue in (b)] showed that not all granule cells are SPCA1 positive [see merge in (b)]. Scale bar: 10 μm.

Table 2 summarizes the SPCA1 distribution in the cerebellar cortex of pig cerebellum with respect to the distribution of the Golgi marker γ adaptin and the other types of Ca2+ pumps.

Table 2.   Summary of the distribution of secretory pathway Ca2+-ATPase 1 (SPCA1), γ adaptin, and other Ca2+ pumps in the pig cerebellar cortex
 SPCA1γ adaptinSERCA2PMCA
  1. SERCA2, sarco(endo)plasmic reticulum Ca2+-ATPase 2; PMCA, plasma membrane Ca2+-ATPase; +, proteins expressed; −, proteins not expressed.

Purkinje cellsSoma Dendritic arborization+ −+ −+ ++ +
Interneurons of molecular layerSoma Dendritic arborization+ −+ −+ Sepúlveda et al. (2004) −−−
Granule cells++++
Cerebellar glomeruli++++

Discussion

The presence of SPCA1 in microsomes and synaptosomes is consistent, respectively, with a broad distribution of the protein in compartments of Golgi complex and in further downstream components of the secretory pathways in neurons. The expression of SPCA as a functional enzyme, as assayed via its corresponding Ca2+-ATPase and 45Ca2+-transport activities, reflects a significant contribution of this pump to the Ca2+-dependent ATP hydrolysis and Ca2+ removal from the cytoplasm in cerebellar neurons. On the other hand, the ability of SPCA to interact with Mn2+, indirectly assessed through the Mn2+ inhibition of 45Ca2+ transport in cerebellar membranes, represents a specific property of this pump not shared with the other Ca2+-ATPases. The fact that increasing concentrations of Mn2+ ions resulted in a corresponding decrease of SPCA Ca2+ uptake and that this Mn2+ inhibition decreased with increasing Ca2+ concentration points to a competition of both cations for the same binding site, as previously suggested by Mandal et al. (2000) for PMR1, the SPCA orthologue in yeast.

The distribution of SPCA1 in sections of pig cerebellum, as definite dots spread over the cytoplasm of the soma of Purkinje cells and interneurons corresponds to the known distribution of the major compartments of the Golgi complex in these cerebellar neurons (Palay and Chan-Palay 1974). This is in good agreement with the presence of SPCA1 in the microsomal fraction, which is enriched in Golgi complex membranes. The SPCA1 localization shown in this study is similar to that reporter by Murin et al. (2006) in neurons and different types of glia in primary cultures derived from the whole rat brain. However, in pig cerebellar cortex, SPCA1 expression seems to be more restricted to neuronal cells. In other non-neuronal cells SPCA1 is found in a compact perinuclear distribution, which corresponds to the location of the Golgi apparatus (Van Baelen et al. 2003; Wootton et al. 2004).

The presence of relatively high levels of SPCA1 in these distinct compartments of neurons might be related to a specific role of the protein for loading Ca2+ or Mn2+ into those reservoirs, but also to the regulation of cytosolic Ca2+ levels controlling synaptic transmission. It has been reported that within the Golgi, Ca2+ is essential for membrane trafficking between the cisternae (Porat and Elazar 2000). SPCA also plays an important role in the conformational maturation and glycosylation of secretory proteins (Durr et al. 1998), and of several permanently or temporarily Golgi-resident membrane proteins. Several of the latter are Ca2+ dependent and known to be implicated in neurodegenerative diseases (Steiner 1998; LaFerla 2002). Thus, SPCA1 might play a pivotal role in maintaining cytosolic as well as organellar (particularly Golgi and the secretory pathway) homeostasis, and alterations in its function could contribute to the etiology of neurological disorders.

The extensive immunoreaction with anti-γ adaptin in the soma of the same neurons where SPCA1 is expressed reflects the presence in their cytoplasm of numerous vesicles derived from the Golgi. This distribution is also consistent with the presence of the SPCA1 protein in synaptosomal fractions, which contain the more distal components of the secretory pathway. A more distal SPCA1 localization in the trans-Golgi network has also been reported for β pancreatic cells from rodents (Mitchell et al. 2004). Moreover, the expression of SPCA1 in cerebellar glomeruli, which contain synaptic terminals lacking Golgi cisternae, suggests a location of SPCA1 in post-Golgi components of the secretory pathway. Although resolution constraints of our immunofluorescence assays prevent us from assigning a more specific localization of SPCA1 within the glomeruli, its mere presence suggests that it could potentially contribute, along with SERCA and PMCA, which are highly expressed in these areas (Sepúlveda et al. 2004), to the local regulation of Ca2+ levels and hence help to control in neurotransmission in these areas with high concentration of synapses. The localization of SPCA1 in cerebellar glomeruli seems to be very specific because SPCA1 was not found in the dendritic arborisation of Purkinje cells, where many synaptic contacts are established and which are enriched in SERCA and PMCA pumps (Sepúlveda et al. 2004).

Little is known about the physiological role of SPCA as a Mn2+ transporter. It has been reported that cytosolic Mn2+ can become toxic for cells, and even elicit apoptosis, most likely by interfering with the Mg2+ binding sites of many proteins and thereby affecting their normal activity (Beckman et al. 1985; Towler et al. 2000; Hirata 2002). In this respect SPCA1 may play a key role in cellular detoxification of Mn2+ ions, similar to what has been described for PMR1 in yeast. PMR1 contributes to the removal of Mn2+ excess from the yeast cytoplasm, by first accumulating Mn2+ into the Golgi compartment and its subsequent elimination from the cell by exocytosis (Lapinskas et al. 1995; Wei et al. 1999; Mandal et al. 2000). In nervous system, Mn2+ exposure can cause neurotoxicity and a neurologic syndrome of Parkinsonism because high levels of Mn2+ inhibit tyrosine hydroxylation, affecting dopamine synthesis and causing the disorder (Olanow 2004). This disorder is very pronounced in people intoxicated by inhalation of manganese powders or vapors, like in steel factory workers or welders, etc. The extent to which SPCA1-related alterations in Ca2+ and Mn2+ homeostasis contribute to the etiology of different types of neurological disorders remains to be further explored.

Acknowledgements

We thank Dr M. Ramirez for generously lending us the microscope and Dr J. A. Armengol for stimulating discussion with the immunohistochemical assays. This work was supported by Grant BFU2005-00663 (to AM Mata) from Ministerio de Educación y Ciencia, Spain. MR Sepúlveda received a visiting postdoctoral fellowship from the FWO Vlaanderen GP.027.07N, Belgium. D Marcos is a recipient of a PhD studentship FPI from Junta de Extremadura, Spain.

Ancillary