Organelles Containing Inositol Trisphosphate Receptor Type 2 in ...

J. Physiol. Sci. Vol. 56, No. 6; Dec. 2006; pp. 415–423 Online Nov. 3, 2006; doi:10.2170/physiolsci.RP006406

REGULAR PAPER

Organelles Containing Inositol Trisphosphate Receptor Type 2 in Adrenal Medullary Cells Yutaka ENDO1, Keita HARADA1, Naoji FUJISHIRO1, Hisasachi FUNAHASHI2, Seiji SHIODA2, Glenn D. PRESTWICH3, Katsuhiko MIKOSHIBA4, and Masumi INOUE1 1Department of Cell and System Physiology, University of Occupational and Environmental Health School of Medicine, Kitakyushu,

807-8555 Japan; 2Department of Anatomy, Showa University School of Medicine, Tokyo, 142-8555 Japan; 3Department of Medical Chemistry, University of Utah, 419 Wakara Way, Suite 205, Salt Lake City, UT 84108, USA; and 4Division of Molecular Neurobiology, The Institute of Medical Science, The University of Tokyo, Tokyo, 108-8639 Japan

Abstract: To identify which organelles contained inositol trisphosphate (InsP3) receptor type 2 (InsP3R2) in adrenal medullary (AM) cells, immunocytochemical and biochemical studies were performed on AM cells of several species. InsP3R2-like immunoreactive materials produced by two different anti-InsP3R2 antibodies (Abs) (Chemicon and Sigma) were distributed in rat AM cells in agreement with BODIPY-FL-InsP3 binding sites. For two other Abs (KM1083 and Santa Cruz), some of the antiInsP3R2 immunoreactive materials were stained with an antidopamine-β-hydroxylase Ab, but not by BODIPY-FL-InsP3. BODIPY-FL-thapsigargin binding sites were consistent with a distri-

bution of the endoplasmic reticulum (ER) identified by an anticalnexin Ab, and a prior application of thapsigargin significantly eliminated BODIPY-FL-thapsigargin bindings, suggesting that BODIPY-FL-thapsigargin bindings were mediated by thapsigargin, but not the fluorescence molecule. The anti-InsP3R2 Ab that produced stainings consistent with BODIPY-FL-InsP3 bindings recognized a protein with about 250 kDa. A fractional analysis of bovine adrenal medullae revealed that the 250 kDa InsP3R2 was detected in a crude membrane fraction, but not in a secretory granule fraction. The results suggest that the InsP3R2 was present in the ER, but not in secretory granules in AM cells.

Key words: inositol trisphosphate receptor, ER, secretory granule, chromaffin cell.

I

t is generally believed that the endoplasmic reticulum (ER) functions as an intracellular Ca2+ store site [1–4]. In goblet cells, however, secretory granules were reported to release Ca2+ in response to inositol 1,4,5-trisphosphate (InsP3) [5], whereas in submandibular acinar cells InsP3 receptors (InsP3Rs) were shown to be present in vesicles, but not in the rough ER [6]. Moreover, Yoo and his colleagues reported that all three subtypes of InsP3Rs resided in secretory granules as well as the ER in bovine adrenal medullary (AM) cells [7–9]. From a functional point of view, the presence of InsP3Rs in secretory granules is readily explained. If InsP3Rs are located in secretory granules underlying the plasma membrane, a stimulation of receptors associated with phospholipase C would induce an increase in [Ca2+] near the plasma membrane. This would efficiently facilitate the fusion of the secretory granule membrane with the plasma membrane and result in secretion [10]. However, secretory granules are organelles totally different from the ER, having a unique set of proteins [11, 12]. Furthermore, the pH in the secretory granules is more acidic than that in the ER [13]. If InsP3Rs are present

in secretory granules, the cell will require special transport and anchoring systems for the granules to maintain them [14]. Our immunocytochemical study [15] using an antibody against calnexin, an integral protein of the ER membrane, revealed that in rat AM cells the ER is distributed not only in the vicinity of the nucleus, but also underneath the plasma membrane. Therefore, InsP3Rs below the plasma membrane are not necessarily located in secretory granules, but they are possibly present in the peripheral ER [16]. In pancreatic acinar cells, InsP3Rs in the apical pole were reported to be present in secretory granules [17, 18]. However, later studies suggested that the InsP3-induced Ca2+ release from a crude granule fraction might have come from the ER contaminated in the fraction [19, 20]. These results led us to investigate the notion that the InsP3Rs at the cell periphery of AM cells may not be present in secretory granules. In the present experiment, we examine this possibility by combining immunocytochemistry with biochemical methods.

Received on Jun 7, 2006; accepted on Nov 1, 2006; released online on Nov 3, 2006; doi:10.2170/physiolsci.RP006406 Correspondence should be addressed to Masumi Inoue, Department of Cell and System Physiology, University of Occupational and Environmental Health School of Medicine, Kitakyushu, 807-8555 Japan. Tel: +81-93-691-7235, Fax: +81-93-602-9883, E-mail: [email protected]

The Journal of Physiological Sciences Vol. 56, No. 6, 2006

415

Y. ENDO et al. METHODS

Immunocytochemistry. Unless specified, immunocytochemistry was performed in dissociated rat AM cells. Male rats and guinea pigs were killed by cervical dislocation, and the adrenal glands were excised. Adrenal medullae were cut into three to six pieces and incubated for 30 min with 0.25% collagenase dissolved in Ca2+-deficient saline (137 mM NaCl, 5.4 mM KCl, 0.5 mM MgCl2, 0.53 mM NaH2PO4, 5 mM D-glucose, 5 mM Hepes, and 4 mM NaOH, pH 7.4). After incubation, the tissues were washed in Ca2+-deficient saline, and placed in a dish, then dissociated using fine needles. The dissociated AM cells were fixed in 2% paraformaldehyde-containing phosphatebuffered saline (PBS: 145 mM NaCl, 8 mM Na2HPO4, 2 mM NaH2PO4) for 2 h at 4°C, then treated in 5% goat or rabbit serum containing PBS with 0.3% Triton X for 30 min. For indirect immunofluorescence studies, the cells were treated with a rabbit anti-InsP3R type 2 (InsP3R2) Ab produced by Sigma (at a dilution of 1:100) overnight, rabbit anti-InsP3R2 Ab by Chemicon (1:10) overnight, goat anti-InsP3R2 Ab by Santa Cruz (1:100) for one or two days, or mouse anti-InsP3R2 Ab KM1083 [21] (1:1,000) overnight, and/or with a rabbit anti-dopamine-β-hydroxylase (DβH) Ab (Chemicon) (1:500) overnight. The AM cells were also treated with a rabbit anti-calnexin Ab (Stressgen) (1:400) overnight. The immunoreactions were visualized with respective secondary Abs conjugated with Alexa 488 or 546 (Molecular Probe) (1:300). In the case of Alexa 488 (FITC-like fluorescence), immunoreactivity was observed with an illumination of a 488 nm laser and an emission of 510–525 nm, whereas in the case of Alexa 546 (rhodamine-like fluorescence), it was observed with an illumination of a 543 nm laser and emission above 560 nm. Cross-talk between FITC- and rhodamine-like fluorescence in double stainings was assessed in experiments where FITC- or rhodamine-like fluorescence was observed after treatment with one secondary Ab coupled with Alexa 546 or 488, respectively. In our experimental conditions, there was no significant cross-talk between FITC- and rhodamine-like fluorescence. To identify organelles stained with the anti-InsP3R2 Abs, cells with immunoreactions were immersed in PBS containing 30 µM BODIPY-FL-InsP3 [16], 1 µM ER Tracker (Molecular Probe), or 4.4 µM 3,3´-dihexyloxacarbocyanine iodide (DIO) (Molecular Probe) whereas to study the relation between the ER and Ca2+ store sites, the cells treated with the anti-calnexin Ab were placed in 0.5 µM BODIPY-FLthapsigargin (Molecular Probe). The immunoreactivity was observed using a confocal laser scanning microscope (Zeiss LSM 410) with X63 objective lenses (a numerical aperture of 1.25 or 1.4). The binding of ligands conjugated with BODIPY moiety (fluorescent ligands) and DIO was visualized with an illumination of a 488 nm laser and an emission of 510–525 nm, whereas ER Tracker bindings 416

were visible with an illumination of 365 nm and emission above 397 nm. Fluorescence was observed through the whole cell with a full width at a half-maximum of 0.7 µm, and the colocalization of InsP3R2-like immunoreactivity with fluorescent agent binding was analyzed at a section where both fluorescences were strongest. Whole-cell images were acquired with an illumination of the 488 nm laser and an emission of all wavelengths, and the cell shape or edge was determined by looking at staining intensity above background levels. A superimposition of cell shape and fluorescence images was made with the use of Photoshop software (version 6.0). To examine the specificity for the immunoreaction, the preparation was treated with a nonimmune serum instead of a primary Ab, and almost no immunoreaction was observed under the same conditions as those used for the primary Ab. Immunoblot. Rat adrenal medullae were homogenized in a Tris buffer (10 mM Tris-HCl [pH 7.4], 150 mM NaCl) containing a protease inhibitor cocktail (Calbiochem). Homogenates were spun down at 2,000 × g for 10 min at 4°C, and the postnuclear supernatants were then mixed with equal volumes of a twofold concentrated Laemmli sample buffer. The crude membrane fraction was obtained with a centrifugation of the postnuclear supernatant at 100,000 × g for 1 h. The resulting pellet was dissolved in the sample buffer. Since catecholamines interfere with a BCA protein assay kit (Pierce), protein concentrations in samples were measured after the proteins were precipitated by acetone and resuspended in the sample buffer. Just before electrophoresis, 5% (v/v) 2-mercaptoethanol and 1% (w/v) bromophenol blue were added to the sample, and the proteins were separated by 10% or 7.5% (w/v) SDS-PAGE, then transferred to a PVDF membrane. The membrane was blocked with 5% (w/v) fat-free powdered milk dissolved in the Tris buffer with 0.1% Tween 20 and then incubated with primary Abs. The immunoreaction was detected by incubating the membrane with respective secondary Abs linked to horseradish peroxidase and then with ECL-Plus (Amersham). Fractionation study. Bovine secretory granules were isolated, as described previously [22]. Briefly, bovine adrenal medullae were homogenized in SME buffer (300 mM sucrose, 10 mM MOPS, and 5 mM EDTA), filtered through one layer of fine mesh cotton gauge. Following centrifugation at 1,000 × g for 15 min, the postnuclear supernatant was spun at 25,000 × g for 20 min to obtain a crude granule pellet and a postgranule supernatant. The pellet was resuspended in SME buffer and layered over step gradients of Percoll (60%, 40%, and 20%). The gradients were spun at 10,000 × g for 30 min. The material banded at the 40% to 60% interface was resuspended in SME buffer, then centrifugated at 25,000 × g for 20 min. This procedure was twice repeated to remove Percoll. The resulting pellet was dissolved in the sample buffer, resulting in a secretory granule fraction. On the other hand, the


Estrogen and Muscle Atrophy

postgranule supernatant was spun at 100,000 × g for 1 h, and the resulting pellet was dissolved in the sample buffer, the solution was designated a crude membrane fraction. Immunogold electron microscopy. Female guinea-pigs were deeply anesthetized with sodium pentobarbital (50 mg/kg, i.p.) and perfused through the ascending aorta with 50 ml of saline, then with 250–300 ml of PBS containing 4% paraformaldehyde and 0.1% glutaraldehyde for 30 min. The adrenal glands were removed immediately and postfixed in the same fixative for 1 day at 4°C. After being washed with PBS, the fixed tissues were trimmed to smaller pieces containing the medulla, incubated for 2 days in 30% sucrose in PBS, and incubated for 3 days in cold PBS with 1.84 M sucrose and 20% polyvinyl pyrrolidone. The tissues were cut into very small pieces and rapidly frozen in liquid nitrogen. Frozen ultrathin sections (150 nm thick) were made with a Lica Ultracut S/FCS (Austria). The sections were picked up on a formvar/carbon-coated nickel grid, then incubated overnight with the Santa Cruz anti-InsP3R2 Ab diluted at 1:100 in the PBS with 2% gelatin and 10 mM glycine. The sections were washed five times with PBS containing 0.5% gelatin and 10 mM glycine, then incubated for 2 h with 10 nm colloidal gold-labeled rabbit anti-goat IgG Ab (2 µg/ml) (British Biocell International). After being washed again, the sections were postfixed in 0.1% glutaraldehyde, stained with 2% uranyl acetate, embedded in polyvinyl alcohol, and observed with a Jeol JEM-1200 EXII electron microscope (Japan).

RESULTS

Immunocytochemistry for InsP3Rs

We previously showed that rat AM cells had mainly the InsP3R2 isoform [23]. Thus to identify intracellular organelles having InsP3R2, an immunostaining of InsP3R2 was compared with the binding of BODIPY-FL-InsP3, which was shown to represent the ER [15]. Figure 1 shows immunostainings obtained with anti-InsP3R2 Abs produced by Chemicon, Sigma, and Santa Cruz (SC) and with KM1083, monoclonal anti-InsP3R2 Ab. The Chemicon anti-InsP3R2 Ab-immunoreactive materials, which were observed as rhodamine-like fluorescence, were distributed both in the vicinity of the nucleus and at the cell periphery in dissociated AM cells (Fig. 1A). In all 24 cells examined, this distribution of immunoreactivity was nearly coincident with the binding of BODIPY-InsP3 (Fig. 1B), which was visible as FITC-like fluorescence. On the other hand, the Sigma Ab produced not only immunoreactive materials in the vicinity of the nucleus and the plasma membrane, but also dotlike materials in the nucleus (Fig. 1C). The distribution of the former generally agreed with that of fluorescent InsP3 bindings in 13 cells, whereas the latter lacked fluorescent InsP3 bindings (Fig. 1D). The distribution of materials immunoreactive to KM1083 that were present in a punctate fashion in the vicinity of the nucleus and the plasma membrane nearly overlapped with BODIPY-FL-InsP3 binding in 2 of 9 cells examined, whereas in the remaining 7, strips at the cell periphery were additionally stained with KM1083, and this

Fig. 1. Immunostaining of rat adreChemicon Ab Fluo IP3 Sigma Ab Fluo IP3 nal medullary cells for InsP3R2. A, A B C D C, E, and G represent confocal images of adrenal medullary (AM) cells stained with an anti-InsP3R2 Ab produced by Chemicon, one by Sigma, KM1083, and one by Santa Cruz (SC), respectively, whereas B, D, F, and H are those of BODIPY-FL-InsP3 bindings in the same level as A, C, E, and G, reKM1083 Fluo IP3 SC Ab Fluo IP3 spectively. AM cells were treated for one or two days with antiE F G H InsP3R2 Abs, then incubated with an Alexa 546-conjugated secondary Ab. After a washout of the secondary Ab, the cells were immersed in 30 µM BODIPY-FLInsP3-containing PBS. Immunoreactions and fluorescent InsP3 bindings were observed as rhodamine5 Pm and FITC-like fluorescence, respectively. The sites at the cell periphery indicated by the closed arrowheads were stained with both Abs and fluorescent InsP3. The open arrowheads in C indicate nonspecific immunoreactions. The strip indicated by the arrow was immunoreactive to KM1083, but lacked fluorescent InsP3 binding. Asterisks represent the nucleus in this and the following figures. The calibration of 5 µm applies from A to H.

*

*

*

* *


417

Y. ENDO et al. Fig. 2. Presence of InsP3R2-like immuER Tracker KM1083 noreactivity in the endoplasmic reticulum A B near the nucleus. A and B represent confocal images of rat AM cells stained with ER Tracker and KM1083 at the same level, respectively. ER Tracker binding and immunoreaction to KM1083 were visualized with an illumination of a 365 nm laser and emission above 397 nm and with an illumination of a 488 nm and emission of 510–525 nm, respectively. C and D show confocal imagDIO es of a rat AM cell stained with DIO and the C D Sigma Ab E Sigma anti-InsP3R2 Ab, respectively. DIO binding and immunoreaction to the Sigma Ab were visible as FITC- and rhodaminelike fluorescence, respectively. The images were a superimposition of cell profile (gray) and DIO binding (blue) or immunoreaction to the Sigma Ab (green) (see METHODS for details). E represents a superimposition of DIO G KM1083 H D and sites (red) with both DIO binding and F immunoreaction to the Sigma Ab. F and G show confocal images of a rat AM cell stained with DIO and KM1083 at the same level. DIO binding and immunoreaction to 5 Pm KM1083 were visible as FITC- and rhodamine-like fluorescence, respectively. The images were a superimposition of cell profile (gray) and DIO binding (blue) or immunoreaction to KM1083 (green). H represents a superimposition of G and sites (red) with both DIO binding and immunoreaction to KM1083. The sites indicated by the arrowheads were stained with both DIO and KM1083, whereas the strip indicated by the arrow was labeled by KM1083, but not by DIO. The calibration of 5 µm applies from A to H.

*

*

*

*

*

striplike immunoreactivity was devoid of fluorescent InsP3 binding (Fig. 1, E and F). On the other hand, the SC anti-InsP3R2 Ab produced immunostaining patterns that were entirely different from BODIPY-FL-InsP3 bindings in 12 cells (Fig. 1, G and H). The materials immunoreactive to the Ab were mainly confined to the cell periphery, which showed no fluorescent InsP3 binding. Taken together, the materials immunoreactive to so-called antiInsP3R2 Abs consisted of two components: one set of materials distributed in a punctate fashion coincided with fluorescent InsP3 bindings in the cytoplasm, whereas the other set present as a strip at the cell periphery lacked them. ER markers and anti-InsP3R2-like immunoreactive materials

Since InsP3R2 is generally thought to be present in the ER, we performed double stainings with anti-InsP3R2 Abs and ER markers, ER Tracker, and DIO [24]. The distribution of KM1083-immunoreactive materials in the vicinity of the nucleus agreed with that of ER Tracker bindings in 11 cells (Fig. 2, A and B). Similarly, materials immunoreactive to the Sigma Ab in the cytoplasm coincided with DIO bindings in 7 cells, whereas immunoreactive materials present in the nucleus had no DIO binding activity (Fig. 2, C, D, and E). These results indicate that 418

materials immunoreactive to anti-InsP3R2 Abs, whose distribution agrees with BODIPY-FL-InsP3 bindings, reside in the ER. InsP3R2-like immunoreactive materials and secretory granules

Our observation that striplike immunoreactive materials to KM1083 at the cell periphery lacked fluorescent InsP3-binding activity raised the possibility that they were located in organelles other than the ER. This possibility was confirmed in Fig. 2, F, G, and H, which revealed that the KM1083-immunoreactive strip along the cell periphery was not stained with DIO either. This finding was consistently observed in 7 cells examined. These two lines of evidence indicate that part of KM1083-immunoractive materials present at the cell periphery were located in organelles other than the ER, most probably secretory granules. This possibility was investigated with a double staining. As shown in Fig. 3, A, B, and C, the region labeled by an Ab raised against DβH that resides in secretory granules [25] overlapped minimally to that of fluorescent InsP3 binding in 8 cells, suggesting that the anti-DβH Ab had not labeled the ER. Most of the KM1083-immunoreactive area along the cell periphery was stained with the anti-DβH Ab in 7 cells (Fig. 3, D, E, and F). Similarly,


Estrogen and Muscle Atrophy Fig. 3. Materials immunoractive to the Santa Cruz Ab and KM1083 are present in secretory granules. A and B show confocal images of a rat AM cell stained with an anti-DβH Ab and BODIPY-FL-InsP3, respecKM1083 DEH tively. C represents a superimposition of A and B, and fluorescent InsP3 binding sites with DβHlike immunoreactivity (IR) were expressed in red. D and E show confocal images of rat AM SC Ab cells stained with the DEH anti-DβH Ab and KM1083, respectively. F represents a superimpo1 Pm sition of E and sites (brown) with immunoreactions to both the Abs. G and H show confocal images of a guinea pig 5 Pm AM cell with the anti-DβH Ab and SC anti-InsP3R2 Ab, respectively. I represents a superimposition of H and sites (brown) with immunoreactions to both the Abs. J shows an immunoelectron micrograph of guinea-pig AM cells treated with the SC anti-InsP3R2 Ab (see METHODS for details). Arrowheads indicate deposits of gold in secretory granules. The insertion is 3× enlargement of the square indicated. The calibration of 5 µm applies from A to I.

A

DEH

B

Fluo IP3

C

D

E

F

G

H

I

J

Fig. 4. Immunoblots of rat adrenal Chemicon KM1083 Santa Cruz medullae with various anti-InsP3R2 Abs. Immunoblots of homogenate H S C H S C H S C (H), supernatant (S), and crude membrane fraction (C) of rat adre250 250 250 nal medulla for InsP3R2. The PVDF 150 150 150 membranes were treated overnight 100 100 100 with the Chemicon Ab (at a dilution 75 75 75 of 1:100), KM1083 (1:4,000), and 50 Santa Cruz Ab (1:2,000). The same 50 50 amounts of proteins (5 and 15 µg) 37 were loaded in each fraction. The 37 37 exposure time for the detection of kDa kDa kDa bands was adjusted so that the InsP3R2 levels detected with the Chemicon Ab were comparable to those with KM1083. These immunoblots represent 3 to 4 experiments for the Abs. The postnuclear homogenate (H) was spun at 100,000 × g for 1 h, and the resulting supernatant and pellet were designated as supernatant (S) and a crude membrane fraction (C) (see METHODS for details).

most of the SC Ab-immunoreactive area was stained with the anti-DβH Ab in 7 cells (Fig. 3, G, H, and I). These immunocytochemical findings indicated that most strips stained with KM 1083 and the SC Ab at the cell periphery were secretory granules. This notion was further confirmed at the electron microscopic level. Figure 3 J clearly revealed that immunoreaction with the SC Ab selectively deposited gold particles in secretory granules.

Immunoblot

The foregoing results raised the possibility that the ER and secretory granules were immunoreactive to putative anti-InsP3R2 Abs, as shown in bovine AM cells [9]. However, this complexity might be ascribed to heterogeneous immunoreactivities of the Abs used. Thus we investigated the specificity of Abs using immunoblotting. The Chemicon Ab specifically recognized a protein of about 250 kDa


419

Y. ENDO et al.

Intensity (au)

Fig. 5. The distribution of thapsigargin bindThap + Fluo Thap Calnexin-like IR ing sites agrees with that of the endoplasmic P < 0.001 A B C reticulum recognized by the anti-calnexin an100 tibody. A and B represent confocal images of 80 a rat AM cell stained with 0.5 µM BODIPY-FLthapsigargin in the presence of 10 µM thapsi60 gargin and with an anti-calnexin Ab at the 40 same level, respectively. After treatment with the anti-calnexin Ab and then a secondary 20 Ab, the cell was exposed to 0.5 µM BODIPYFL-thapsigargin in the presence of 10 µM + thapsigargin. BODIPY fluorescence and calthapsigargin Fluo Thap Calnexin-like IR nexin-like IR were visible as FITC- and rhodamine-like fluorescence, respectively. C D E summarizes the fluorescence levels of BODIPY-FL-thapsigargin bindings in the presence and absence of thapsigargin. The fluorescence intensity of BODIPY-FLthapsigargin bindings was expressed in arbitrary units (au). The means and s.e.m. are shown (see RESULTS for details). The statistical significance was determined with a Student’s t-test. D and E represent confocal images of a rat AM cell stained with BODIPYFL-thapsigargin (Fluo Thap) and anti-calnexin Ab at the same level, respectively. BODIPY fluorescence and calnexin-like IR were visible as FITC- and rhodamine-like fluorescence, respectively.

in a homogenate and a crude membrane fraction of the rat adrenal medulla and not in a supernatant fraction (Fig. 4), indicating that the Ab was specific for InsP3R2. On the other hand, the SC Ab did not detect the 250 kDa protein, although it recognized bands with lower molecular weights. KM1083 mainly recognized the 250 kDa protein, but it also recognized bands with lower molecular weights in the homogenate, supernatant, and crude membrane fraction. The results indicated that the Chemicon Ab was most selective for InsP3R2 among the three Abs, at least in the rat. Distribution of fluorescent thapsigargin bindings

A

B

DEH HO

C

SG

CM

HO SG

CM

250 150 100 75

250 150 100 75

50

50

37 kDa

37 kDa

IP3R2 HO

SG

CM

Our previous study indicated that Ca2+ store sites sensi-

250

tive to muscarinic receptor stimulation in the rat AM cells were depleted by thapsigargin [23], a specific inhibitor of sarcoplasmic and/or endoplasmic reticulum Ca2+ (SERCA) pumps [26], and Ca2+ mobilization in response to muscarinic receptor stimulation in AM cells was most likely mediated by InsP3 [27, 28]. These results suggest that SERCA pumps are responsible for Ca2+ uptake into the InsP3-sensitive Ca2+ store sites. SERCA pumps are firmly established to be present in the SR in the muscles [2–4], but in other cells [24, 29] they were also very likely to reside in other organelles, such as the Golgi. Thus we studied whether SERCA pumps were located in the ER. To explore the distribution of SERCA pumps, rat AM cells were exposed to BODIPY-FL-thapsigargin, and the binding was visualized as FITC-like fluorescence. Figure 5 shows that BODIPY-FL-thapsigargin binding was not due to the fluorescent moiety, but to thapsigargin itself.

150

420

Calnexin

100 75 kDa Fig. 6. Fractional analysis of InsP3R2 distribution in bovine adrenal medullae. A and B represent immunoblots of homogenate (HO), highly purified secretory granule (SG), and crude membrane fraction (CM) for DβH and calnexin, respectively. Note that SG and CM fractions were rich in DβH and calnexin contents, respectively. C shows an immunoblot of HO, SG, and CM for InsP3R2. InsP3R2 was detected in HO and CM, but not in SG with the Chemicon Ab. In A and B, 25 µg proteins were loaded in each lane, whereas in C 15 µg were loaded.

When fluorescent intensities of FITC-like fluorescence at 3 to 9 sites per cell, which corresponded to those with cal-


Estrogen and Muscle Atrophy

nexin-like immunoreactivity visible as rhodamine-like fluorescence, were measured in 6 cells with and without a 10 min prior exposure to 20 µM thapsigargin (Fig. 5, A and B), mean intensity per pixel (pixel size = 0.079 µm × 0.079 µm) was significantly diminished by the prior exposure (Fig. 5C). In 13 cells, the distribution of fluorescent thapsigargin binding (Fig. 5, D and E) overlapped with that of the ER, as identified by immunostaining for calnexin, an integral protein of the ER. Fractionation study

The presence of InsP3R2 in secretory granules was further studied by the fractionation of a postnuclear homogenate of the bovine adrenal medulla. A highly purified secretory granule (SG) fraction, which was obtained in two steps using centrifugation with a Percoll gradient, was rich in DβH and had almost no calnexin, whereas a crude membrane (CM) fraction, which was obtained with centrifugation of a postgranular supernatant, was rich in calnexin, but lacked DβH (Fig. 6, A and B). Figure 6 C reveals that InsP3R2 was detected in the CM fraction, but not in the SG fraction, suggesting that InsP3R2 is present in the ER, but not in secretory granules. DISCUSSION

Secretory granules, such as chromaffin [30, 31] and zymogen granules [32] in pancreatic acinar cells, are known to contain large amounts of Ca2+ ions. The physiological significance of these intragranular Ca2+ ions ([Ca2+]g), although suggested to be involved in secretion [7], still remains obscure. Chromaffin and zymogen granules were shown to release Ca2+ ions in response to InsP3 [8, 17], and InsP3Rs were localized in chromaffin granules by the use of immunogold electron microscopy [9]. Consistent with the notion that secretory granules are dynamic Ca2+ stores, the measurement of [Ca2+] in insulin-containing granules of MIN6 cells with a vesicle-associated membrane protein aequorin chimera revealed a dramatic decrease in [Ca2+]g in response to caffeine and cyclic ADP ribose [11]. However, the use of a similar approach to measure [Ca2+]g in neurosecretory PC12 cells revealed that [Ca2+]g was altered with a change in intracellular pH [33], but it did not noticeably respond to InsP3 or cyclic ADP ribose [12]. Furthermore, the presence of InsP3Rs in zymogen granules has been disputed [19, 20]. In the present experiment, four different Abs were used to elucidate organelles having InsP3R2 in rat AM cells. The Sigma Ab was produced in rabbits with the synthetic peptide corresponding to amino acid residues 317–334 of the rat InsP3R2. The materials immunoreactive to the Ab in the vicinity of the nucleus and cell membrane bound to the same regions as BODIPY-FL-InsP3 did. A similar result was obtained with the Chemicon Ab, which was made in rabbits with a peptide corresponding to the carboxy ter-

minus of the rat InsP3R2. The BODIPY-FL-InsP3 binding sites below the plasma membrane as well as in the vicinity of the nucleus in chromaffin cells were previously shown to be identical with sites stained with an anti-calnexin Ab [15], whereas materials immunoreactive to the Sigma Ab were distributed similarly to DIO binding sites, except for the nucleus. These previous and present results indicate that central and peripheral sites with InsP3R2-like immunoreacitivity that agree with BODIPY-FL-InsP3-binding sites represented the ER. On the other hand, the SC antiInsP3R2 Ab, which was raised in goats against a peptide mapping at the carboxy terminal of human InsP3R2, stained strips along the cell periphery, which were labeled with an anti-DβH Ab, but not BODIPY-FL-InsP3, indicating that the immunoreactive materials to the SC Ab were present in secretory granules. This notion was confirmed at the electron microscopic level. The immunoreactive properties of KM1083, which was also raised against the carboxy terminal of human InsP3R2 [21], were between those of the rabbit Abs and the goat Ab. Most of these mAb-labeled materials in the vicinity of the cell membrane were stained with the anti-DβH Ab, but not with BODIPY-FL-InsP3, whereas the mAb-sensitive area near the nucleus was stained with the latter and ER Tracker, which selectively bind to the ER. In summary, the four Abs could be divided into two groups: one , including the Sigma and Chemicon Abs, did not stain strips at the cell periphery; the other, comprising the SC Ab and KM1083, did stain them. The immunocytochemical studies with the various Abs in the present and previous experiments [23] might suggest that InsP3R2 resides not only in the ER, but also in secretory granules, as proposed in bovine AM cells [7]. However, we think that this possibility is not feasible. First, the Chemicon Ab specifically recognized InsP3R2 of about 250 kDa in rat adrenal medullae. On the other hand, KM1083 in the second group mainly recognized the 250 kDa protein, but it also recognized proteins with lower molecular weights whereas the SC Ab scarcely detected the InsP3R2. Thus the Chemicon Ab appeared to be most specific for the InsP3R2 at least in the rat. This Ab showed a staining similar to BODIPY-FL-InsP3-binding sites in the vicinity of the nucleus and cell membrane, but it produced no striplike stainings at the cell periphery. Our notion that the InsP3R2 is present in the ER, but not in secretory granules, was further supported by fractionation studies revealing that the 250 kDa InsP3R2 was recovered in a CM fraction, but not in a highly purified SG fraction. Furthermore, BODIPY-FL-thapsigargin -binding sites coincided with those labeled with the anti-calnexin Ab. We [23] and others [34] reported that thapsigargin depleted Ca2+ ions from store sites, from which Ca2+ mobilization occurred in response to InsP3-generating receptor agonists. Thus the coincidence of BODIPY-FL-thapsigargin-binding sites with calnexin-like immunoreactive sites


421

Y. ENDO et al.

is consistent with the notion that the ER alone is responsible for InsP3-sensitive store sites in chromaffin cells. These immunocytochemical and biochemical studies suggest that InsP3R2 is distributed in the ER, but not in secretory granules in rat adrenal chromaffin cells. This notion compares with electron microscopic studies of the cerebellum, which revealed that immunoreactive products to an anti-InsP3R1 Ab were present in the ER, but not in other organelles, such as the Golgi complex or synaptic vesicles [35, 36]. Yoo’s group recently extended immunogold electron microscopic studies to show that InsP3R-like immunoreactive materials were distributed not only in the ER and secretory granules, but also in the plasma membrane, mitochondria, and nuclear envelope [37]. Furthermore, the nucleoplasm was also shown to have InsP3R-like products, as was noted with the Sigma anti-InsP3R2 Ab. Thus according to a series of papers by Yoo and his colleagues [9, 37], InsP3Rs turn out to be present almost everywhere in bovine chromaffin cells. This apparent complexity of InsP3R distributions, in our opinion, may at least in part be due to nonspecific binding. There was a similar controversy about whether InsP3Rs were present in secretory granules in pancreatic β cells or not. The InsP3R type 3 was reported to be present in insulin-containing secretory granules [38, 39]. However, its presence in insulin granules was contradicted because the Ab that was used (AB3Ab) cross-reacted with rat insulin [40]. Thus to elucidate which organelles contain target proteins, we would need to perform immunocytochemical studies with multiple Abs, which were raised against different epitopes. This work was supported in part by a Grant-in-Aid from the Japan Society for the Promotion of Science (13670050 to M. I.) and by a grant from the National Institutes of Health (NS29632 to G. D. P.). Thanks are also due to S Ozaki for preparing the BODIPY-FL-InsP3 and to Hai Lin for a preliminary experiment. REFERENCES 1. Berridge MJ. Inositol trisphosphate and calcium signalling. Nature. 1993;361: 315-25. 2. Blaustein MP, Golovina VA. Structural complexity and functional diversity of endoplasmic reticulum Ca2+ stores. Trends Neurosci. 2001;24:602-8. 3. Meldolesi J, Pozzan T. The endoplasmic reticulum Ca2+ store: a view from the lumen. Trends Biochem Sci. 1998;23:10-4. 4. Pozzan T, Rizzuto R, Volpe P, Meldolesi J. Molecular and cellular physiology of intracellular calcium stores. Physiol Rev. 1994;74:595-636. 5. Nguyen T, Chin W-C, Verdugo P. Role of Ca2+/K+ ion exchange in intracellular storage and release of Ca2+. Nature. 1998;395:908-12. 6. Yamamoto-Hino M, Miyawaki A, Segawa A, Adachi E, Yamashina S, Fujimoto T, Sugiyama T, Furuichi T, Hasegawa M, Mikoshiba K. Apical vesicles bearing inositol 1,4,5-trisphosphate receptors in the Ca2+ initiation site of ductal epithelium of submandibular gland. J Cell Biol.1998;141:135-42. 7. Yoo SH. Coupling of the IP3 receptor/Ca2+ channel with Ca2+ storage proteins chromogranins A and B in secretory granules. Trends Neurosci. 2000;23:424-8. 8. Yoo SH, Albanesi JP. Inositol 1,4,5-trisphosphate-triggered Ca2+ release from bovine adrenal medullary secretory vesicles. J Biol Chem.1990;265:13446-8. 9. Yoo SH, Oh YS, Kang MK, Huh YH, So SH, Park HS, Park HY. Localization of three types of the inositol 1,4,5-trisphosphate receptor/ Ca2+ channel in the secretory granules and coupling with the Ca2+ storage proteins chromogranins A and B. J Biol Chem. 2001;276:45806-12.

422

10. Tse FW, Tse A, Hille B, Horstmann H, Almers W. Local Ca2+ release from internal stores controls exocytosis in pituitary gonadotrophs. Neuron. 1997;18:121-32. 11. Mitchell KJ, Pinton P, Varadi A, Tacchetti C, Ainscow EK, Pozzan T, Rizzuto R, Rutter GA. Dense core secretory vesicles revealed as a dynamic Ca2+ store in neuroendocrine cells with a vesicle-associated membrane protein aequorin chimaera. J Cell Biol. 2001;155:41-51. 12. Pouli AE, Karagenc N, Wasmeier C, Hutton JC, Bright N, Arden S, Schofield JG, Rutter GA. A phogrin-aequorin chimaera to image free Ca2+ in the vicinity of secretory granules. Biochem J. 1998;330:1399-404. 13. Schoonderwoert VThG, Martens GJM. Proton pumping in the secretory pathway. J Membr Biol. 2001;182:159-69. 14. Parker AKT, Gergely FV, Taylor CW. Targeting of inositol 1,4,5-trisphosphate receptors to the endoplasmic reticulum by multiple signals within their transmembrane domains. J Biol Chem. 2004;279:23797-805. 15. Lin H, Ozaki S, Fujishiro N, Takeda K, Imanaga I, Prestwich GD, Inoue M. Subunit composition and role of Na+,K+-ATPases in adrenal chromaffin cells. J Physiol. 2005;564:161-72. 16. Inoue M, Sakamoto Y, Fujishiro N, Imanaga I, Ozaki S, Prestwich GD, Warashina A. Homogeneous Ca2+ stores in rat adrenal chromaffin cells. Cell Calcium. 2003; 33:19-26. 17. Gerasimenko OV, Gerasimenko JV, Belan PV, Petersen OH. Inositol trisphosphate and cyclic ADP-ribose-mediated release of Ca2+ from single isolated pancreatic zymogen granules. Cell. 1996;84:473-80. 18. Petersen OH. Can Ca2+ be released from secretory granules or synaptic vesicles?. Trends Neurosci. 1996;19:411-3. 19. Nezu A, Tanimura A, Morita T, Irie K, Yajima T, Tojyo Y. Evidence that zymogen granules do not function as an intracellular Ca2+ store for the generation of the Ca2+ signal in rat parotid acinar cells. Biochem J. 2002;363:59-66. 20. Yule DI, Ernst SA, Ohnishi H, Wojcikiewicz RJH. Evidence that zymogen granules are not a physiologically relevant calcium pool. Defining the distribution of inositol 1,4,5-trisphosphate receptors in pancreatic acinar cells. J Biol Chem. 1997;272: 9093-8. 21. Sugiyama T, Furuya A, Monkawa T, Yamamoto-Hino M, Satoh S, Ohmori K, Miyawaki A, Hanai N, Mikoshiba K, Hasegawa M. Monoclonal antibodies distinctively recognizing the subtypes of inositol 1,4,5-trisphosphate receptor: application to the studies on inflammatory cells. FEBS Lett. 1994;354:149-54. 22. Meyer DI, Burger MM. Isolation of a protein from the plasma membrane of adrenal medulla which binds to secretory vesicles. J Biol Chem. 1979;254:98549. 23. Inoue M, Lin H, Imanaga I, Ogawa K, Warashina A. InsP3 receptor type 2 and oscillatory and monophasic Ca2+ transients in rat adrenal chromaffin cells. Cell Calcium. 2004;35:59-70. 24. Hauser K, Pavlovic N, Klauke N, Geissinger D, Plattner H. Green fluorescent protein-tagged sarco(endo)plasmic reticulum Ca2+-ATPase overexpression in Paramecium cells: isoforms, subcellular localization, biogenesis of cortical calcium stores and functional aspects. Mol Microbiol. 2000;37:773-87. 25. Winkler H, Apps DK, Fiseher-Colbrie R. The molecular function of adrenal chromaffin granules: established facts and unresolved topics. Neuroscience. 1986;18:261-90. 26. Lytton J, Westlin M, Hanley MR. Thapsigargin inhibits the sarcoplasmic or endoplasmic reticulum Ca-ATPase family of calcium pumps, J Biol Chem.1991; 266:17067-71. 27. Inoue M, Sakamoto Y, Imanaga I. Phosphatidylinositol hydrolysis is involved in production of Ca2+-dependent currents, but not non-selective cation currents, by muscarine in chromaffin cells. Eur J Pharmacol. 1995;276:123-9. 28. Ohta T, Asano T, Ito S, Kitamura N, Nakazato Y. Characteristics of cytosolic Ca2+ elevation induced by muscarinic receptor activation in single adrenal chromaffin cells of the guinea pig. Cell Calcium. 1996;20:303-14. 29. Lin P, Yao Y, Hofmeister R, Tsien RY, Farquhar MG. Overexpression of CALNUC (Nucleobindin) increases agonist and thapsigargin releasable Ca2+ storage in the Golgi. J Cell Biol. 1999;145:279-89. 30. Fasolato C, Zottini M, Clementi E, Zacchetti D, Meldolesi J, Pozzan T. Intracellular Ca2+ pools in PC12 cells. Three intracellular pools are distinguished by their turnover and mechanisms of Ca2+ accumulation, storage, and release. J Biol Chem. 1991;266:20159-67. 31. Winkler H, Westhead E. The molecular organization of adrenal chromaffin granules. Neuroscience. 1980;5:1803-23. 32. Clemente F, Meldolesi J. Calcium and pancreatic secretion. I. Subcellular distribution of calcium and magnesium in the exocrine pancreas of the guinea pig. J Cell Biol. 1975;65:88-102. 33. Moreno A, Lobatón CD, Santodomingo J, Vay L, Hernández-SanMiguel E, Rizzuto R, Montero M, Alvarez J. Calcium dynamics in catecholamine-


Estrogen and Muscle Atrophy containing secretory vesicles. Cell Calcium. 2005;37:555-64. 34. Poulsen JCJ, Caspersen C, Mathiasen D, East JM, Tunwell REA, Lai FA, Maeda N, Mikoshiba K, Treiman M. Thapsigargin-sensitive Ca2+-ATPases account for Ca2+ uptake to inositol 1,4,5-trisphosphate-sensitive and caffeine-sensitive Ca2+ stores in adrenal chromaffin cells. Biochem J. 1995;307:749-58. 35. Satoh T, Ross CA, Villa A, Supattapone S, Pozzan T, Snyder SH, Meldolesi J. The inositol 1,4,5,-trisphosphate receptor in cerebellar Purkinje cells: quantitative immunogold labeling reveals concentration in an ER subcompartment. J Cell Biol. 1990;111:615-24. 36. Volpe P, Villa A, Damiani E, Sharp AH, Podini P, Snyder SH, Meldolesi J. Heterogeneity of microsomal Ca2+ stores in chicken Purkinje neurons. EMBO J. 1991;10:3183-9. 37. Huh YH, Yoo JA, Bahk SJ, Yoo SH. Distribution profile of inositol 1,4,5-

trisphosphate receptor isoforms in adrenal chromaffin cells. FEBS Lett. 2005;579:2597-603. 38. Blondel O, Bell GI, Moody M, Miller RJ, Gibbons SJ. Creation of an inositol 1,4,5trisphosphate-sensitive Ca2+ store in secretory granules of insulin-producing cells. J Biol Chem. 1994;269:27167-70. 39. Blondel O, Moody MM, Depaoli AM, Sharp AH, Ross CA, Swift H, Bell GI. Localization of inositol trisphosphate receptor subtype 3 to insulin and somatostatin secretory granules and regulation of expression in islets and insulinoma cells. Proc Natl Acad Sci USA. 1994;91:7777-81. 40. Ravazzola M, Halban PA, Orci L. Inositol 1,4,5-trisphosphate receptor subtype 3 in pancreatic islet cell secretory granules revisited. Proc Natl Acad Sci USA. 1996;93: 2745-8.


423