Evaluation, using targeted aequorins, of the roles of the endoplasmic

Transcription

1 This is the Pre-Published Version Evaluation, using targeted aequorins, of the roles of the endoplasmic reticulum and its (Ca 2+ + Mg 2+ )ATP-ases in the activation of storeoperated Ca 2+ channels in liver cells Caroline Chan, 2 M. Lyn Harland, 1 Sarah E. Webb, 2 Jinglong Chen, 1 Andrew L. Miller 2 and Greg J. Barritt 1 1 Department of Medical Biochemistry, School of Medicine, Faculty of Health Sciences, Flinders University, G.P.O. Box 2100, Adelaide, South Australia, 5001, Australia; and 2 Department of Biology, The Hong Kong University of Science & Technology, Clear Water Bay, Kowloon, Hong Kong, China Address for Correspondence and Proofs: Professor G.J. Barritt, Department of Medical Biochemistry, School of Medicine, Faculty of Health Sciences, Flinders University, G.P.O. Box 2100, Adelaide, South Australia, 5001, Australia Telephone: (+61 8) Fax: (+61 8) Greg.Barritt@flinders.edu.au

2 -2- Summary The process by which store-operated Ca 2+ channels (SOCs) deliver Ca 2+ to the endoplasmic reticulum (ER) and the role of (Ca 2+ + Mg 2+ )ATP-ases of the ER in the activation of SOCs in H4-IIE liver cells were investigated using cell lines stably transfected with apo-aequorin targeted to the cytoplasmic space or the ER. In order to measure the concentration of Ca 2+ in the ER ([Ca 2+ ] er ), cells were pre-treated with 2,5- di-tert-butylhydroquinone (DBHQ) to deplete Ca 2+ in the ER before reconstitution of holo-aequorin. The addition of extracellular Ca 2+ (Ca 2+ o) to Ca 2+ -depleted cells induced refilling of the ER, which was complete within 5 min. This was associated with a sharp transient increase in the cytoplasmic Ca 2+ concentration ([Ca 2+ ] cyt ) of about 15 s duration (a Ca 2+ o-induced [Ca 2+ ] cyt spike) after which [Ca 2+ ] cyt remained elevated slightly above the basal value for a period of about 2 min (low [Ca 2+ ] cyt plateau). The Ca 2+ o-induced [Ca 2+ ] cyt spike was: inhibited by Gd 3+, not affected by tetrakis-(2-pyridymethyl) ethylenediamine (TPEN), and broadened by ionomycin and the intracellular Ca 2+ chelators BAPTA and EGTA. Refilling of the ER was inhibited by caffeine. Neither thapsigargin nor DBHQ caused a detectable inhibition or change in shape of the Ca 2+ o-induced [Ca 2+ ] cyt spike or the low [Ca 2+ ] cyt plateau whereas each inhibited the inflow of Ca 2+ to the ER by about 80%. Experiments conducted with carbonyl cyanide m-chlorophenyl-hydrazone (CCCP) indicated that thapsigargin did not alter the amount of Ca 2+ accumulated in mitochondria. The changes in [Ca 2+ ] cyt reported by aequorin were compared with those reported by fura-2. It is concluded that (i) there are significant quantitative differences between the manner in which aequorin and fura-2 sense changes in [Ca 2+ ] cyt and (ii) thapsigargin and DBHQ inhibit the uptake of Ca 2+ to the bulk of the ER but this is not associated with inhibition of the activation of SOCs. The possible involvement of a small sub-region of the ER (or

3 -3- another intracellular Ca 2+ store), which contains thapsigargin-insensitive (Ca 2+ + Mg 2+ )ATP-ases, in the activation of SOCs is briefly discussed. Abbreviations: ER, endoplasmic reticulum; SERCA, sarcoplasmic reticulum (Ca 2+ + Mg 2+ )ATP-ase; cytaeq, aequorin targeted to the cytoplasmic space; eraeq, aequorin targeted to the ER; [Ca 2+ ] cyt, the concentration of free Ca 2+ in the cytoplasmic space; [Ca 2+ ] er, the concentration of free Ca 2+ in the ER; Ca 2+ o, extracellular Ca 2+ ; KRB, Krebs-Ringer bicarbonate buffer; DBHQ, 2,5-di-tert-butylhydroquinone; BAPTA, 1,2- bis(2-aminophenoxy)ethane-n,n,n,n -tetraacetic acid; CCCP, carbonyl cyanide m- chlorophenyl-hydrazone; PBS, phosphate buffered saline; HA1, haemaglutinin; TPEN, tetrakis-(2-pyridymethyl) ethylenediamine.

4 -4- INTRODUCTION Store-operated Ca 2+ channels (SOCs) are present in most non-excitable animal cells [1-3]. Their physiological role is thought to be to replenish the endoplasmic reticulum (ER) Ca 2+ stores during and after the action of agonists which induce the release of Ca 2+ from the ER through InsP 3 and ryanodine receptors [1-3]. The activation of SOCs is initiated by a decrease in Ca 2+ in the ER [1-3]. Under physiological conditions, this decrease is caused by the actions of InsP 3 and Ca 2+ in the cytoplasmic space. However, it can be achieved artificially by inhibitors of the ER (Ca 2+ + Mg 2+ )ATP-ase (SERCA) such as thapsigargin and 2,5-di-tert-butylhydroquinone (DBHQ) [1,3]. There is evidence for several sub-types of SOCs, which can be distinguished on the basis of their different selectivities for Ca 2+ compared with Na + [3]. The SOCs most extensively characterised are the Ca 2+ release-activated Ca 2+ channels (CRAC channels) present in mast cells and lymphocytes [4]. One feature of CRAC channels is their strong inhibition by increased concentrations of Ca 2+ at the cytoplasmic mouth of the channel [4,5]. Ca 2+ inflow through SOCs in hepatocytes and liver cell lines has chiefly been characterised using fluorescent Ca 2+ sensors and patch-clamp recording techniques [6,7]. Recent studies with H4-IIE cells (a liver cell line derived from the rat Reuber hepatoma) have shown that the principal type of SOC has the same high Ca 2+ selectivity and other properties as CRAC channels. It was concluded that the properties of the principal SOCs in H4-IIE cells are indistinguishable from those of the CRAC channels in mast cells and lymphocytes [7].

5 -5- The mechanism of activation of SOCs has not yet been elucidated, although numerous experiments have been conducted and several hypotheses proposed [1-4,8]. Some experimental results indicate that pre-requisites for the activation of SOCs are (i) the location of regions of the ER close to the plasma membrane, (ii) the maintenance of the normal integrity of the actin cytoskeleton and ER [8,9] and (iii) the normal function of SERCAs [10-13]. There is also some evidence that the activation of SOCs and flow of Ca 2+ through the channels involves the movement of Ca 2+ from the extracellular space to the ER through a small region of the cytoplasmic space, the subplasmalemmal space, without diffusion to the bulk of the cytoplasmic space (reviewed in [14]) (shown schematically in Fig. 1). Other experiments suggest that the activation of SOCs may require the fusion of vesicles containing SOC proteins with the plasma membrane (reviewed in [2,8]). It is presently hypothesised that the mechanism of SOC activation involves a small molecule such as Ca 2+ influx factor, conformation coupling, and/or insertion of vesicles containing SOCs into the plasma membrane [8], or that activation is achieved simply by a decrease in the concentration of Ca 2+ in the subplasmalemmal space, which de-inhibits the Ca 2+ -inhibitable SOCs [14]. The aim of the present experiments was to obtain a better understanding of (i) the relationship between the concentration of Ca 2+ in the cytoplasmic space ([Ca 2+ ] cyt ) and the concentration of Ca 2+ in the lumen of the ER ([Ca 2+ ] er ) during the activation of SOCs and refilling of the ER, and (ii) the role of SERCAs in the activation mechanism. The experiments reported have been performed with H4-IIE liver cells using aequorin targeted to the cytoplasmic space (cytaeq) or lumen of the ER (eraeq) to measure changes in Ca 2+ in these intracellular locations. The results

6 -6- indicate that there are significant quantitative differences between the changes in [Ca 2+ ] cyt reported by aequorin, on the one hand, and fluorescent probes on the other. These may be due, in part, to differences in their Ca 2+ buffering, in their affinity, and in their dynamic range of response to Ca 2+. It was also found that the changes in [Ca 2+ ] cyt detected by aequorin located in the cytoplasmic space (a transient spike and subsequent low plateau) associated with re-filling of the ER (i.e. when Ca 2+ o is added to cells in which the ER has been depleted of Ca 2+ ) are insensitive to inhibition by thapsigargin or DBHQ, whereas these agents inhibit the inflow of Ca 2+ to the bulk of the ER. These observations suggest the possibility that a small sub-region of the ER which possesses thapsigargin- and DBHQ-insensitive SERCAs is responsible for the activation of SOCs in H4-IIE liver cells.

7 -7- MATERIALS AND METHODS Materials Reagents were obtained from the following sources: Dulbecco s Modified Eagle s Medium (DMEM), penicillin, streptomycin, foetal calf serum (FCS), lipofectamine 2000 and G418 from Life Technologies Inc (GIBCO BRL) Rockville, MD, USA; monoclonal antibody 12CA5 from Roche Diagnostics Australia Pty. Ltd., Castle Hill, NSW, Australia; Cy3-conjugated donkey anti-mouse IgG secondary antibody from Jackson ImmunoResearch Lab Inc., Baltimore, USA; rabbit anti-calnexin antibody from Stressgen Biotechnologies Corp, San Diego, CA, USA; FITC-conjugated antirabbit IgG from Zymed Laboratories Inc; San Francisco, USA; DBHQ from Sapphire Bioscience Pty Ltd, Crows Nest, NSW, Australia; thapsigargin, caffeine, ATP, ionomycin, 2-aminoethoxydiphenylborate (2-APB), GdCl 3, collagen and dimethylsulphoxide from Sigma; coelenterazine and coelenterazine derivatives (f and n), tetrakis-(2-pyridymethyl) ethylenediamine (TPEN) and the acetoxymethylesters of fura-2, 5,5 dibromo BAPTA, 5,5'-dimethyl BAPTA, EGTA, eraeq/pcdna1 and cytaeq/pcdna1 from Molecular Probes. Transfection of H4-IIE liver cells and selection of stable clones expressing cytaeq and eraeq H4-IIE rat hepatoma cells (ATCC CRL 1548) were maintained in DMEM, supplemented with penicillin (100 units/ml), streptomycin (100 µg/ml), 10 mm

8 -8- HEPES (ph 7.4) and 10% (v/v) foetal calf serum (complete DMEM), as described previously [6,15]. H4-IIE cells were subcultured for a maximum of 20 passages. cdna encoding a cytoplasmic space-targeted apo-aequorin (cytaeq/pcdna1) was comprised of cdna encoding the HA1 (haemaglutinin) epitope-tagged apo-aequorin (cytaeq) in the pcdna1 vector. cdna encoding an ER-targeted apo-aequorin (eraeq/pcdna1) was comprised of cdna encoding a peptide leader sequence fused to HA1 epitope-tagged low Ca 2+ affinity mutated apo-aequorin (eraeq) in the pcdna1 expression vector [16]. cytaeq/pcdna1 and eraeq/pcdna1 were subcloned into the pcdna3 vector containing the neomycin gene neo. Transfection of H4-IIE cells with eraeq/pcdna3 and cytaeq/pcdna3 was carried out using Lipofectamine Cells (1 x 10 5 ) were plated onto a 35 mm tissue culture dish 24 h before transfection. DNA (6 µg diluted in 200 µl DMEM) was mixed with Lipofectamine 2000 (12 µl diluted in 200 µl DMEM) incubated for 20 min, then added to the cells and incubated for a further 24 h. Two days after transfection, selection was started with 0.8 mg/ml G418. Single colonies were transferred to a 24 well plate, expanded and tested for the presence of DNA encoding aequorin using PCR. Positive clones harbouring cdna encoding apo-eraeq or apo-cytaeq were identified, and the levels of aequorin luminescence in each determined. Clones cyt241 (cytaeq) and er334 (eraeq) which exhibited the highest luminescence were used in the present experiments. The clones were maintained in the presence of G418 (0.4 mg/ml).

9 -9- Immunolocalization of expressed eraeq and cytaeq To determine the intracellular localisation of cytaeq and eraeq, H4-IIE cells stably expressing cytaeq or eraeq were fixed for 30 min with 4% (v/v) paraformaldehyde in phosphate buffered saline (PBS) (2.9 mm KCl, 137 mm NaCl, 1.5 mm KH 2 PO 4, 9.5 mm Na 2 HPO 4 ). They were then washed twice with PBS and permeabilised with 0.1% (v/v) Triton-X 100 in PBS for 4 min on ice followed by a 45 min incubation in 15% (w/v) FCS in PBS. The cells were then incubated for 30 min with the monoclonal antibody 12CA5 which recognizes the HA1 epitope tag, at a dilution of 1:100. Staining was carried out by incubation for 30 min with the Cy3-conjugated donkey anti-mouse IgG secondary antibody at a dilution of 1:100. After each antibody incubation, cells were washed 4-5 times with PBS. For double immunostaining of both eraeq and calnexin in cells expressing eraeq, cells were incubated for 30 min with a mixture of 12CA5 (at a dilution of 1:100) and rabbit anti-calnexin antibody (at a dilution of 1:200) followed by incubation with a mixture of Cy3-conjugated donkey anti-mouse IgG (at a dilution of 1:100) and FITC-conjugated anti-rabbit IgG (at a dilution of 1:100). The cells were then mounted under 80% glycerol in PBS, ph 8.6 and examined under a Leica DMRBE fluorescence microscope (Leica Microsystems, Heidelberg, Germany) using a Leica PL FLUOTAR 100x/1.30 NA objective, and excitation filters 575 nm (Cy3) and 490 nm (FITC). Fluorescent images were captured with the LEICA DC 100 software. Cell nuclei were identified using Hoechst [17].

10 -10- Measurement of aequorin luminescence Glass tubes (12 x 75 mm) (Falcon) were coated with collagen (500 µg/ml rat tail collagen type 1) for 20 min, then UV sterilized. H4-IIE cells, suspended in complete DMEM, were transferred into collagen-coated tubes at 1 x 10 5 cells per tube and incubated for 24 h (eraeq transfected cells) or 2 h (cytaeq transfected cells) at 37 o C in a humidified incubator (5% (v/v) CO 2 in air) in order to establish a cell monolayer on the bottom of the tube. Except where indicated otherwise, one of two protocols was employed to reconstitute holo-aequorin. In the first, the [Ca 2+ ] er was reduced by incubating the monolayer of eraeq or cytaeq cells for 5 min with the SERCA inhibitor, 2,5-di-tert-butylhydroquinone (DBHQ) (25 µm) in Ca 2+ -free Krebs-Ringer modified buffer (KRB; 125 mm NaCl, 5 mm KCl, 1 mm Na 2 HPO4, 1 mm MgSO 4, 5.5 mm glucose and 20 mm HEPES, ph 7.4), supplemented with 3 mm EGTA. The cells were then washed with KRB (in the absence of added Ca 2+ ). Holo-aequorin reconstitution was then carried out by incubating the cells at 22 C for 1 h with either coelenterazine n (eraeq) or coelenterazine (cytaeq) at 5 µm in KRB containing 1 mm EGTA. In the second protocol, the monolayer of cytaeq cells was washed in KRB then holo-aequorin reconstitution was carried out by incubating the cells at 22 C with coelenterazine (5 µm) in KRB containing 1 mm CaCl 2. After holo-aequorin reconstitution, cells were washed once with KRB. Specific subsequent treatments are described in the figure legends. Aequorin luminescence was measured at 22 C using an FB15 luminometer (Zylux corporation, Maryville, TN, USA) fitted with reagent injectors and linked to an IBM-compatible PC and Zylux Corporation FB12 software. Reagents were added manually by a Gilson pipette. At

11 -11- the end of the experiment, the cells were lysed by addition of CaCl 2 (10 mm final concentration) and digitonin (100 µm), both added via the reagent injectors. Stock solutions of thapsigargin, DBHQ, TPEN, 2-APB and the acetoxymethylesters of 5,5 dibromo BAPTA, 5,5'-dimethyl BAPTA and EGTA were prepared in dimethylsulphoxide (DMSO) whilst stock solutions of caffeine, ATP and GdCl 3 were made up in ultra-pure Milli Q water. The final concentration of DMSO was 1% (v/v). For experiments involving Gd 3+, Na 2 HPO 4 was omitted from the KRB to prevent Gd 3+ precipitation. Values of luminescence were converted to [Ca 2+ ] cyt or [Ca 2+ ] er using the calibration curves published by Brini et al. [18] (cytaeq) and Barrero et al. [19] (eraeq). These employed the same cytaeq and eraeq proteins and the same temperature (22 C) as in the present experiments. Values of aequorin luminescence, L (relative luminescence units/s), from the FB15 luminometer were transferred to Microsoft Excel 2000 (Microsoft Corp, Redmond, WA, USA). Total luminescence, L max (relative luminescence units/s), was estimated from the sum of the luminescence observed during each component of the experiment and the luminescence induced when cells were lysed in the presence of 10 mm CaCl 2. The value of L at a given time was converted to L/L max using Microsoft Excel The calculated log L/L max ratio was then imported into the GraphPad PRISM 3 software and analysed by non-linear regression curve fitting using a polynomial second order equation in order to convert values of Log L/L max to [Ca 2+ ] cyt or [Ca 2+ ] er. Values of [Ca 2+ ] cyt or [Ca 2+ ] er were then exported to Microsoft Excel 2000 and values of [Ca 2+ ] cyt or [Ca 2+ ] er plotted as a function of time. Means ± SEM of replicate experiments were calculated and the SEM

12 -12- bars are shown on the graphs at 15 s or minute intervals. Degrees of significance were determined using the Student s t-test for unpaired samples. Imaging of endoplasmic reticulum-targeted aequorin luminescence For photon imaging, eraeq cells were plated onto plastic coverslips (22 x 22 mm; Scienceware, Bel-Art Inc, Pequannock, USA) at approximately 1 x 10 5 cells/coverslip. The coverslips were then mounted underneath a hole cut in a 35 mm petri-dish lid using high vacuum silicone grease (Dow Corning Corp, Midland, MI, USA) and the cells incubated for 24 h at 37 o C in a humidified incubator with 5% (v/v) CO 2 in air in order to create a monolayer of cells on the coverslip. Holo-aequorin was reconstituted with 5 µm coelenterazine f [20] after incubation of the cells with 25 µm DBHQ as described above. (The relative light intensity emitted from holo-aequorin reconstituted from coelenterazine f is about 130 times greater than that emitted from holo-aequorin reconstituted from coelenterazine n [20].) Photons were imaged as described previously [21]. Measurement of the cytoplasmic Ca 2+ concentration using fura-2 and fura-2ff H4-IIE cells were loaded with fura-2 using the acetoxymethyl ester, and the fluorescence ratio (proportional to [Ca 2+ ] cyt ) was measured using 340 and 380 nm excitation filters (Chroma Technology Corp. Vt., USA) as described previously [22]. The proportion of fura-2 in the cytoplasmic space was estimated to be about 70%.

13 -13- RESULTS Intracellular location of cytoplasmic and ER-targeted aequorins Stable H4-IIE cell lines expressing haemaglutinin (HA1)-tagged aequorin targeted to the cytoplasmic space (cytaeq cells) and ER (eraeq cells) were prepared as described in the Materials and Methods. The intracellular locations of eraeq and cytaeq were determined via immunocytochemistry with an anti-ha1 antibody and a Cy3-conjugated secondary antibody. The intracellular distribution of immunofluorescence attributed to cytaeq is shown in Fig. 2A (c.f. Fig. 2B, the conjugate control in which the anti-ha1 antibody was omitted). The results obtained for the cell shown and for other cells examined indicated that cytaeq is chiefly located in the cytoplasmic space. The fluorescence image for the location of ERtargeted aequorin is shown in Fig. 2C (c.f. Fig. 2F, the conjugate control in which the secondary antibody was omitted) and the image of calnexin, an ER-protein [23], in Fig. 2D. As indicated by the merged images shown in Fig. 2E, there is considerable co-localisation of eraeq with calnexin. The conclusion that eraeq is located in the ER is consistent with this result. Thus it is concluded that cytaeq and eraeq are located in the cytoplasmic space and ER, respectively. Characterisation of the ability of cytaeq to sense changes in the cytoplasmic free Ca 2+ concentration To characterise the ability of cytaeq to sense changes in [Ca 2+ ] cyt, DBHQ was added to cytaeq cells incubated in the presence of extracellular Ca 2+ (Ca 2+ o) to release Ca 2+

14 -14- from the ER and induce Ca 2+ inflow. A peak of increased [Ca 2+ ] cyt with a tail of about 2 min duration was observed (Fig. 3A, KRB, Ca 2+ ). When the experiment was performed in the absence of added Ca 2+ o, the resulting plot was not significantly different from that obtained in the presence of 2 mm Ca 2+ o (Fig. 3A, KRB), suggesting that no increase in [Ca 2+ ] cyt (manifest as a plateau after the initial DBHQ-induced increase in [Ca 2+ ] cyt ) due to Ca 2+ inflow could be detected. To further test whether any component of the DBHQ-induced increase in [Ca 2+ ] cyt is due to Ca 2+ inflow through SOCs, the effects of Gd 3+, an inhibitor of SOCs [6], were investigated. When added in either the presence or absence of Ca 2+ o, Gd 3+ (10 µm) had no significant effect on the shape of the curve of DBHQ-induced increase in [Ca 2+ ] cyt (Figs. 3B and 3C). Characterisation of changes in [Ca 2+ ] cyt associated with refilling of the ER Ca 2+ store To measure Ca 2+ in the ER, the protocol developed previously by others [16,19,24] was followed. This involved treating H4-IIE cells with DBHQ to release Ca 2+ from the ER, reconstituting holo-aequorin with coelenterazine n in the absence of added Ca 2+ o, incubating cells initially in the absence of added Ca 2+ o, then adding Ca 2+ o. Following Ca 2+ o addition to eraeq cells, [Ca 2+ ] er increased and reached a plateau after about 4 min (Fig. 4A). The curve stops abruptly because at this time point all aequorin had reacted, as indicated by the absence of any further increase in luminescence when the cells are subsequently lysed in the presence of Ca 2+ (results not shown). This time course is similar to that observed for refilling of the ER in other cell types [19]. The increase in [Ca 2+ ] er was completely inhibited by Gd 3+ (Fig. 4A).

15 -15- When changes in [Ca 2+ ] cyt were monitored in cytaeq cells pre-treated with DBHQ, the addition of Ca 2+ o was found to induce a very sharp spike of increased [Ca 2+ ] cyt (Ca 2+ o-induced [Ca 2+ ] cyt spike) with a maximum value of about 10 µm followed by a tail (low [Ca 2+ ] cyt plateau) in which [Ca 2+ ] cyt remained very slightly elevated above the basal value for about 30 sec (Fig. 4B). The duration of the spike was about 15 s. Subsequent addition of digitonin (to lyse the cells) and 10 mm Ca 2+ (to react remaining aequorin) caused a substantial increase in luminescence (not shown). It was calculated that about 85% of the aequorin remained unreacted after Ca 2+ o addition. The duration of the Ca 2+ -induced [Ca 2+ ] cyt spike (Fig. 4B) was noticeably shorter than that induced by DBHQ addition in the presence of Ca 2+ o (Fig. 3A). To determine whether the characteristics of the Ca 2+ o-induced [Ca 2+ ] cyt spike were affected by the concentration of Ca 2+ o employed, the Ca 2+ add-back was performed with two sequential additions of Ca 2+ o, 1 and 10 mm (Fig. 4C). For both 1 and 10 mm Ca 2+ o additions, after about sec [Ca 2+ ] cyt returned to the same base level as that before Ca 2+ o addition. Similar results were obtained when 1 mm Ca 2+ o was added in place of 10 mm Ca 2+ o (not shown). The spike of increased [Ca 2+ ] cyt induced by Ca 2+ o addition was substantially reduced in the presence of Gd 3+ (Fig. 4D). Correlation of the increases in [Ca 2+ ] cyt with increases in [Ca 2+ ] er indicates that the Ca 2+ o-induced [Ca 2+ ] cyt spike occurs in the early stages of ER refilling. [Ca 2+ ] cyt returned to a value just above the basal level when [Ca 2+ ] er was only about 20% refilled, and remained slightly elevated while refilling of the ER continued. The ability of 2-APB, another inhibitor of plasma membrane Ca 2+ channels (reviewed in [25,26]) to inhibit the Ca 2+ o-induced [Ca 2+ ] cyt spike and refilling of the

16 -16- ER was also tested. While 2-APB addition did inhibit Ca 2+ o-induced increases in aequorin luminescence, it also inhibited photon release from aequorin in lysed cells (results not shown). This suggests that 2-APB inhibits the reaction catalysed by aequorin which leads to photon release and hence cannot be used in experiments involving aequorin as Ca 2+ sensor. The Ca 2+ o-induced increase in luminescence of eraeq was imaged using a photon imaging microscope. Increased luminescence was observed in most regions of the cell following the addition of Ca 2+ o (Fig. 5). Since the focal plane of the objective lens is very broad, photons arising from above and below the nucleus are detected in this experiment, and hence it is not possible to see the location of the nucleus. Moreover, there is not sufficient resolution to indicate from which part of the ER the luminescence first originates following Ca 2+ o addition. It was noted that the shape of the spike of Ca 2+ o-induced [Ca 2+ ] cyt observed upon addition of Ca 2+ o to cells previously incubated in the absence of added Ca 2+ o (Fig. 4B) is considerably different to the shape of the [Ca 2+ ] cyt spike observed in similar experiments using fura-2 and fluo-3 as a Ca 2+ sensor in H4-IIE cells [6, 27]. In order to try to understand reasons for this difference, the profiles of Ca 2+ o-induced increases in [Ca 2+ ] cyt and [Ca 2+ ] er were further investigated using ATP, a purinergic agonist which has been shown to initiate the generation of InsP 3 in H4-IIE cells [28]; caffeine, which activates ryanodine receptors in the ER [29]; the Ca 2+ ionophore ionomycin; and the Ca 2+ chelators BAPTA and EGTA. Addition of ATP in the absence of Ca 2+ o caused a very small increase in [Ca 2+ ] cyt (Fig. 6A inset). This small ATP-induced increase in [Ca 2+ ] cyt is consistent with the protocol employed to reconstitute aequorin

17 -17- (which leaves the ER Ca 2+ stores essentially depleted of Ca 2+ ). ATP did not substantially alter either the magnitude or the shape of the Ca 2+ o-induced [Ca 2+ ] cyt spike (Fig. 6A). However, ATP caused greater than 50% inhibition of the rate and extent of Ca 2+ inflow to the ER (Fig. 6B). This effect is likely to be due to the activation of InsP 3 receptors and enhancement of Ca 2+ outflow from the ER. Caffeine substantially reduced the accumulation of Ca 2+ in the ER (Fig. 6C), possibly due to the activation of Ca 2+ outflow through ryanodine receptors [29]. Ionomycin induced a substantial increase in [Ca 2+ ] cyt when added in the absence of Ca 2+ o (Fig. 6D, inset). The ionomycin-induced increase in [Ca 2+ ] cyt was much greater than that induced by ATP (Fig. 6A, inset). It would be expected that in addition to affecting Ca 2+ in the ER, ionomycin would induce the release of Ca 2+ from mitochondria and other non-er intracellular Ca 2+ stores. Ionomycin did not alter the peak of the [Ca 2+ ] cyt spike induced by Ca 2+ o addition but did substantially extend the tail (i.e. slowed the return to basal [Ca 2+ ] cyt ) (Fig. 6D). In experiments employing eraeq, ionomycin (0.1 µm) inhibited the uptake of Ca 2+ to the ER by about 50% (results not shown). Loading cytaeq cells with dimethyl BAPTA resulted in a considerable broadening of the Ca 2+ o-induced [Ca 2+ ] cyt spike (Fig. 7A). In eraeq cells loaded with dibromo- BAPTA, the rate and extent of Ca 2+ o-induced filling of the ER was substantially reduced (Fig. 7B). Loading cytaeq cells with 1 mm EGTA also resulted in considerable broadening of the Ca 2+ o-induced [Ca 2+ ] cyt spike (Fig. 7C).

18 -18- Effects of thapsigargin and DBHQ on Ca 2+ o-induced increases in [Ca 2+ ] cyt and [Ca 2+ ] er The addition of thapsigargin to cells previously subjected to the ER Ca 2+ depletion protocol and incubated in the absence of Ca 2+ o induced a small increase in [Ca 2+ ] cyt (inset, Fig. 8A) (c.f. effect of ATP (Fig. 6A, inset)), indicating that a small amount of Ca 2+ remained in the ER after the DBHQ pre-treatment. Thapsigargin had no detectable effect on either the magnitude or the shape of the Ca 2+ o-induced [Ca 2+ ] cyt spike or on the low [Ca 2+ ] cyt plateau which followed the initial Ca 2+ o-induced [Ca 2+ ] cyt spike (Fig. 8A). Thus, in the presence of thapsigargin, [Ca 2+ ] cyt remained elevated slightly above base-line for about 2 min, then decreased to the basal value. By contrast, thapsigargin caused a substantial inhibition of Ca 2+ o-induced inflow of Ca 2+ to the ER (Fig. 8B). Results similar to those for thapsigargin were obtained with DBHQ (Figs. 8C and 8D). When used in combination, thapsigargin and DBHQ did not affect the Ca 2+ o-induced [Ca 2+ ] cyt spike while almost completely inhibiting Ca 2+ inflow to the ER (results not shown). The lack of effect of thapsigargin or DBHQ on the Ca 2+ o-induced [Ca 2+ ] cyt spike is unlikely to be due to either saturation of aequorin with Ca 2+ or exhaustion of the holoaequorin since subsequent addition of digitonin and 10 mm Ca 2+ (to react with remaining aequorin) caused a substantial increase in luminescence (results not shown). The possibility that, in the presence of thapsigargin, Ca 2+ is taken up by mitochondria [30,31] was investigated using CCCP (a mitochondrial uncoupler which releases mitochondrial Ca 2+ [30,31]). The addition of CCCP after Ca 2+ o addition induced a small increase in [Ca 2+ ] cyt, presumably due to the release of Ca 2+ from

19 -19- mitochondria (Fig. 9A). There was no difference in the CCCP-induced increase in [Ca 2+ ] cyt between cells incubated in the presence and absence of thapsigargin (Fig. 9A). The inset of Fig. 9A shows the Ca 2+ o-induced spike of [Ca 2+ ] cyt in each experiment. In another series of experiments, CCCP was added before Ca 2+ o in order to block any Ca 2+ uptake by mitochondria during Ca 2+ o add-back. For cells incubated in either the absence or presence of thapsigargin, the magnitude of the Ca 2+ o-induced [Ca 2+ ] cyt spike was the same whether or not CCCP was present or absent (Fig. 9B). Effects of TPEN on Ca 2+ o-induced increases in [Ca 2+ ] cyt and [Ca 2+ ] er TPEN has been used to chelate Ca 2+ in the ER, decrease [Ca 2+ ] er, and activate SOCs [32]. The ability of TPEN to activate SOCs in H4-IIE cells was tested in the following manner. cytaeq cells (not pre-treated with DBHQ) were incubated with coelenterazine in the presence of 1 mm Ca 2+ o then placed in a medium containing no added Ca 2+ o. Luminescence was then measured as a function of time. This protocol was designed to maintain as much of the Ca 2+ in the ER as possible before the addition of TPEN. In the absence of TPEN, the addition of Ca 2+ o induced a spike of increased [Ca 2+ ] cyt (Fig. 10A). The addition of TPEN to cells treated in this way and incubated initially in the absence of added Ca 2+ o did not substantially affect the Ca 2+ o-induced spike of [Ca 2+ ] cyt (Fig. 10A). Pre-treatment with TPEN almost completely inhibited the signal from eraeq (Fig. 10B), most likely due to the buffering of Ca 2+ in the ER by the TPEN. When TPEN was added to cells incubated in the presence of Ca 2+ o, a spike of increased [Ca 2+ ] cyt was observed and [Ca 2+ ] cyt remained elevated for about 5

20 -20- min (i.e., TPEN induced a low plateau of increased [Ca 2+ ] cyt ) (Fig. 10C). This result suggests that reduction of [Ca 2+ ] er by TPEN can initiate the activation of SOCs. Comparison of changes in the cytoplasmic Ca 2+ concentration reported by cytaeq with those reported by fura-2 Since the results obtained using cytaeq as a sensor for changes in [Ca 2+ ] cyt appear to differ, at least quantitatively, from those reported previously for H4-IIE cells using fura-2 or fluo-3 as a Ca 2+ sensor [6,27], experiments were performed with H4-IIE cells loaded with fura-2 under conditions similar to those employed for the measurement of Ca 2+ using cytaeq. Fura-2 has a higher affinity for Ca 2+ and a lower dynamic range (the concentrations of Ca 2+ over which fura-2 can be used as a Ca 2+ sensor) [33] than aequorin [18, 20, 24, 34]. The addition of DBHQ in the presence of Ca 2+ o induced a prolonged increase in [Ca 2+ ] cyt (monitored using the fura-2 fluorescence ratio) and this was substantially reduced by Gd 3+ (Fig. 11A) or by removal of Ca 2+ o (Fig. 11B (DBHQ addition) c.f. 11A). The addition of Ca 2+ o to fura-2 loaded cells treated with DBHQ in the absence of added Ca 2+ o then washed (in the absence of Ca 2+ o) in order to remove the DBHQ caused a transient increase in fura-2 fluorescence ratio followed by a prolonged plateau (Fig. 11B). A second addition of Ca 2+ o caused no further increase in fluorescence ratio (Fig. 11B). When the experiment was repeated, but without removal of DBHQ (i.e. DBHQ was present during Ca 2+ o addition), the resulting plateau was greater than that observed in the absence of DBHQ (Fig. 11C c.f. 11B). (The magnitude of the Ca 2+ o-

21 -21- induced plateau (after subtraction of the value of fluorescence ratio immediately before Ca 2+ o addition) was 0.15 ± 0.01 and 0.07 ± 0.02 (n=4) for cells loaded with fura-2 and incubated in the presence or absence (after washing) of DBHQ, respectively (P 0.05, Student t-test for unpaired samples). A second addition of Ca 2+ o caused a further small increase in fluorescence ratio (Fig. 11C).

22 -22- DISCUSSION The short transient Ca 2+ o-induced increase in [Ca 2+ ] cyt Three interesting findings reported here are (i) the short duration of the transient increase in [Ca 2+ ] cyt reported by cytaeq, (ii) the apparent difference between the changes in [Ca 2+ ] cyt sensed by cytaeq compared with the changes sensed by the fluorescent Ca 2+ reporter fura-2 following Ca 2+ inflow induced by re-addition of Ca 2+ o, and (iii) the observation that SERCA inhibitors can substantially inhibit the inflow of Ca 2+ to the bulk of the ER yet the activation of SOCs (assessed by the Ca 2+ o-induced increase in [Ca 2+ ] cyt ) is not sensitive to these inhibitors. A particularly striking feature of the Ca 2+ signals reported by cytaeq is the short duration (about 15 sec) of the spike of increased [Ca 2+ ] cyt observed on adding Ca 2+ o to cytaeq cells in which the ER has previously been depleted of Ca 2+. [Ca 2+ ] cyt remained elevated during the period of refilling of the ER (monitored by eraeq) but this elevation in [Ca 2+ ] cyt was very small, especially when compared with the value of [Ca 2+ ] cyt at the peak of the spike. The Ca 2+ o-induced spike of [Ca 2+ ] cyt and low [Ca 2+ ] cyt plateau were substantially reduced by Gd 3+, indicating that the Ca 2+ o-induced spike of [Ca 2+ ] cyt represents Ca 2+ inflow through open SOCs. One explanation for these observations is that, in the absence of SERCA inhibitors, when Ca 2+ o is added to cells in which the ER Ca 2+ has been depleted, Ca 2+ flows through SOCs and is then very rapidly taken up by the ER (shown schematically by the solid line and arrows in Fig. 1). A pseudo steady-state is then reached during refilling of the ER in which there is a small elevation of [Ca 2+ ] cyt (the low [Ca 2+ ] cyt

23 -23- plateau) in all or some of the cytoplasmic space (c.f. [10, 35]). The initial sharp Ca 2+ o- induced spike of increased [Ca 2+ ] cyt may represent the initial inflow of Ca 2+ through open SOCs (the cells were previously incubated with DBHQ to decrease [Ca 2+ ] er and the DBHQ subsequently removed) resulting in a rapid increase in [Ca 2+ ] cyt (possibly in the vicinity of SOCs) then Ca 2+ -dependent feed-back inhibition of SOCs [4, 14] thus creating an overshoot in [Ca 2+ ] cyt. Another possible contributor to the decline in [Ca 2+ ] cyt which occurs after the initial increase is the activity of the plasma membrane (Ca 2+ + Mg 2+ )ATP-ase [36-38]. The Ca 2+ o-induced spike of [Ca 2+ ] cyt was observed when (i) Ca 2+ o was added to cells pre-treated with DBHQ to release Ca 2+ from the ER, (ii) a second addition of Ca 2+ o was made 5 min after the first Ca 2+ o addition to cells in which ER Ca 2+ stores had previously been depleted by treatment with DBHQ, and (iii) Ca 2+ o was added to cells in which holo-aequorin had been reconstituted in the presence of 1 mm Ca 2+ o in the absence of pre-depletion of the stores with DBHQ. Under conditions (ii) and (iii), the ER Ca 2+ stores should be essentially replete. But since there is no evidence that other plasma membrane Ca 2+ channels are open, it is likely that the observed increase in [Ca 2+ ] cyt represents Ca 2+ inflow through SOCs under conditions (ii) and (iii) as well as condition (i). Caffeine, an activator of ryanodine receptors [29], substantially inhibited refilling of the ER. While there is some evidence for the presence of ryanodine receptors in liver cells [39-41], not all workers agree that the hepatocyte ER possesses these receptors [42]. Our results obtained with caffeine provide additional functional evidence for the

24 -24- presence of ryanodine receptors in the ER of liver cells, although the possibility that caffeine activates another type of Ca 2+ channel in the ER [42] cannot be excluded. Comparison of aequorin with fura-2 as a reporter of [Ca 2+ ] cyt Under conditions in which DBHQ was added to cells incubated in the presence or absence of Ca 2+ o and in the presence or absence of Gd 3+, the results obtained using cytaeq as a sensor for changes in [Ca 2+ ] cyt were quantitatively different from those obtained using fura-2. Following the addition of DBHQ in the presence of Ca 2+ o fura- 2 reported a transient increase in fluorescence followed by a substantial sustained increase (c.f. [6, 25, 43]). The latter was inhibited by removal of Ca 2+ o or by addition of Gd 3+ in the presence of Ca 2+ o. The transient increase in fluorescence observed using fura-based Ca 2+ sensors in liver cells and other cell types is attributed to the release of Ca 2+ from the ER while the sustained increase is attributable to the activation of Ca 2+ inflow across the plasma membrane[6, 25, 43]. Thus it appears that, in contrast to fura-2, cytaeq reports a lower increase of [Ca 2+ ] cyt due to enhanced Ca 2+ inflow, as assessed by the plateau after the Ca 2+ o-induced [Ca 2+ ] cyt spike. There are several possible reasons for the differences in changes in [Ca 2+ ] cyt reported by cytaeq and fura-2. These can be briefly summarised as follows. (i) Loading cells with dimethyl BAPTA or EGTA broadened the Ca 2+ o-induced spike of [Ca 2+ ] cyt sensed by cytaeq so that the general shape of the curve describing increases in [Ca 2+ ] cyt as a function of time is more like the plots observed using fura-2. These results suggest that one of the differences may be the additional buffering capacity

25 -25- introduced by fura-2 [44, 45]. Such a buffering by fura-2 of the [Ca 2+ ] cyt signal in cerebellar granule neurones has been proposed to be one explanation for the observation that fura-2 masks a caffeine-induced release of Ca 2+ from the ER [46]. (ii) The presence of some fura-2 in the ER may distort the ability of fura-2 to sense [Ca 2+ ] cyt, although it is considered this is unlikely to be a major reason for differences in [Ca 2+ ] cyt reported by cytaeq and fura-2. (iii) Differences in affinity for Ca 2+ (the apparent dissociation constants for the binding of Ca 2+ to aequorin and fura-2 are about 30 [34] and 0.2 [33] µm, respectively) and dynamic range (fura-2 [33] has a lower dynamic range than aequorin [18,20,24,34]) may account for some of the difference observed with cytaeq and fura-2. (iv) A likely and significant reason is the possibility that under the conditions in which [Ca 2+ ] cyt was measured there are localised regions of high Ca 2+ concentration in the cytoplasmic space. (For example, near the mouth of SOCs during Ca 2+ inflow when [Ca 2+ ] o is added to cells previously incubated in the absence of added [Ca 2+ ] o.) Aequorin can detect localised regions of high Ca 2+ concentration which are not, or may not be, detected by fura-2. Thus, the large dynamic range for Ca 2+ allows aequorin to detect regions of high Ca 2+ concentration (which may be localised and would not be observed with fura-2) but the aequorin in these regions of putative high Ca 2+ concentration is quickly consumed and further aequorin luminescence from those regions depends on the diffusion of fresh aequorin into the regions [37, 47-49]. The steep response curve for the aequorin signal and the differences between aequorin and fura-2 in detecting changes in [Ca 2+ ] cyt reported in the present studies are similar to those described using these two Ca 2+ sensors in studies of arterial smooth muscle cells [48] and for the use of cytaeq as a Ca 2+ sensor in CHO cells [37]. The results

26 -26- described here using cytaeq to measure [Ca 2+ ] cyt (i.e. the Ca 2+ o-induced [Ca 2+ ] cyt spike of a short duration) are similar to those obtained by Woods et al. [50] who microinjected aequorin into freshly-isolated rat hepatocytes in order to monitor changes in [Ca 2+ ] cyt. The present results obtained with aequorin may point to a local region of high Ca 2+ concentration in the cytoplasmic space of H4-IIE cells during Ca 2+ inflow through SOCs. Effects of SERCA inhibitors on the Ca 2+ o-induced increases in [Ca 2+ ] cyt and [Ca 2+ ] er Within the limits of detection, thapsigargin or DBHQ (or a combination of both inhibitors) did not affect either the Ca 2+ o-induced [Ca 2+ ] cyt spike or the low [Ca 2+ ] cyt plateau which followed the spike. However, thapsigargin and DBHQ each separately, or in combination, substantially inhibited the flow of Ca 2+ to the ER as detected by eraeq. The inhibiton was not complete, and there was a small refilling of the ER Ca 2+ store (or of any related intracellular organelles which express and retain eraeq) in the presence of the SERCA inhibitor. Not only was there no inhibition by thapsigargin or DBHQ of the Ca 2+ o-induced spike of increased [Ca 2+ ] cyt and the low [Ca 2+ ] cyt plateau, but the SERCA inhibitors did not increase the magnitude of these parameters. It might have been expected that, in the presence of thapsigargin or DBHQ, Ca 2+ that was prevented from entering the ER would be re-directed to the cytoplasmic space as shown schematically in Figure 1 (broken lines and arrows). Indeed, when fura-2 was used as the Ca 2+ sensor, the plateau of the Ca 2+ o-induced increase in [Ca 2+ ] cyt was higher in the presence of DBHQ than in the absence of this

27 -27- SERCA inhibitor. Similar differences in changes in Ca 2+ concentration reported by cytoplasmic aequorin and fura-2 have been described for arterial smooth muscle cells upon re-addition of Ca 2+ o to cells previously incubated in the absence of added Ca 2+ o in the presence or absence of cyclopiazonic acid [48]. In experiments using cytaeq, the existence of the Ca 2+ o-induced [Ca 2+ ] cyt spike and the low [Ca 2+ ] cyt plateau in the presence of thapsigargin or DBHQ indicates that Ca 2+ can still flow into the cell through open SOCs under these conditions. The results of previous patch clamp recording experiments conducted with H4-IIE cells [7] also clearly indicate that in the presence of thapsigargin (and an intracellular Ca 2+ buffer to prevent Ca 2+ -dependent feedback inhibition) the SOCs remain open and convey Ca 2+. The observation that SOCs are active in the presence of thapsigargin is also consistent with results obtained previously using fluorescent dyes as Ca 2+ sensors in liver cells [6, 27] (and present results) and in many other cell types [1, 3, 14]. Thus in the present experiments, while thapsigargin and DBHQ have inhibited most of the inflow of Ca 2+ to the bulk of the ER, they do not seem to have inhibited the activation of SOCs. But it appears that, after an initial flow of some Ca 2+ through SOCs into the cytoplasmic space, the flow of Ca 2+ stops (most likely due to Ca 2+ - dependent feedback inhibition of the SOCs). Ca 2+ does not flow into the ER (there is no significant refilling of the ER), the mitochondrial (there is no increase in CCCPreleasable Ca 2+ ) or the cytoplasmic space (there is no large increase in [Ca 2+ ] cyt ). The observation that Ca 2+ inflow to the bulk of the ER can be inhibited while the activation of SOCs is not affected suggests (as one of several possible explanations) that a small intracellular Ca 2+ store (rather than the bulk of the ER) is responsible for the activation

28 -28- of SOCs in H4-IIE liver cells. This suggestion is consistent with the results of other experimental approaches in other cell types which have been interpreted to indicate that a small ER Ca 2+ store, which can be functionally distinguished from the bulk of the ER, is required for the activation of SOCs (c.f. [51-53] and references therein). Since the ER is heterogenous [53], and in the present experiments which employ ERtargeted aequorin is defined by the peptide sequence on eraeq which targets aequorin to the ER, it is possible that the small ER Ca 2+ store hypothesised to be involved in the activation of SOCs is another type of intracellular organelle, for example a Golgiderived vesicle (c.f. [54]). Thus, one explanation for the present results is that (i) the activation of SOCs and the maintenance of Ca 2+ inflow through SOCs is achieved by the hypothesised small ER Ca 2+ store [51-54] (ii) (Ca 2+ + Mg 2+ )ATP-ases are required to draw Ca 2+ into this small ER Ca 2+ store and hence to maintain the flow of Ca 2+ through SOCs [10-13], and (iii) the (Ca 2+ + Mg 2+ )ATP-ases involved are insensitive to thapsigargin- and DBHQ [14]. These ideas, which are consistent with the proposals of others on the mechanism of activation of SOCs [51-53], are summarised in Figure 1. However, other explanations are possible, and further experiments are required to test this working hypothesis. These include the use of more refined imaging of changes in [Ca 2+ ] er and the use of patch clamp recording techniques.

29 -29- ACKNOWLEDGEMENTS This work was supported by CERG-RGC grant HKUST6131/99M, RGC-DAG 98/99.SC30, and an Australian Research Council grant awarded to Dr. Andrew L. Miller and Professor Greg J. Barritt. We gratefully acknowledge Mrs. Diana Kassos and Miss Lee-Anne Addis for preparation of the typescript, and advice provided by Dr. Roland Gregory, Department of Medical Biochemistry, Flinders University, and Professor Peter Cobbold, Department of Human Anatomy and Cell Biology, University of Liverpool.

30 -30- REFERENCES 1. Putney JW, Mckay RR. Capacitative calcium entry channels. Bioessays 1999; 21: Elliott AC. Recent developments in non-excitable cell calcium entry. Cell Calcium 2001; 30: Barritt GJ. Receptor-activated Ca 2+ inflow in animal cells: a variety of pathways tailored to meet different intracellular Ca 2+ signalling requirements. Biochem J 1999; 337: Parekh AB, Penner R. Store depletion and calcium influx. Physiol Rev 1997; 77: Zweifach A, Lewis RS. Rapid inactivation of depletion-activated calcium current (I crac ) due to local calcium feed-back. J Gen Physiol 1995; 105: Auld A, Chen J, Brereton HM, Wang Y-J, Gregory RB, Barritt GJ. Storeoperated Ca 2+ inflow in Reuber hepatoma cells is inhibited by voltage-operated Ca 2+ channel antagonists and, in contrast to freshly isolated hepatocytes, does not require a pertussis toxin-sensitive trimeric GTP-binding protein. Biochim Biophys Acta 2000; 1497: Rychkov G, Brereton HM, Harland ML, Barritt GJ. Plasma membrane Ca 2+ release-activated Ca 2+ channels with a high selectivity for Ca 2+ identified by patch-clamp recording in rat liver cells. Hepatology 2001; 33:

31 Zitt C, Halaszovich CR, Lückhoff A. The TRP family of cation channels: probing and advancing the concepts on receptor-activated calcium entry. Prog Neurobiol 2002; 66: Wang YJ, Gregory RB, Barritt GJ. Maintenance of the filamentous actin cytoskeleton is necessary for the activation of store-operated Ca 2+ channels, but not other types of plasma-membrane Ca 2+ channels, in rat hepatocytes. Biochem J 2002; 363: Mogami H, Nakano K, Tepikin AV, Petersen OH. Ca 2+ flow via tunnels in polarized cells: recharging of apical Ca 2+ stores by focal Ca 2+ entry through basal membrane patch. Cell 1997; 88: Liu X, Rojas E, Ambudkar IS. Regulation of K Ca current by store-operated Ca 2+ influx depends on internal Ca 2+ release in HSG Cells. Am J Physiol 1998; 275: C571-C Liu X, O Connell A, Ambudkar IS. Ca 2+ -dependent inactivation of a storeoperated Ca 2+ current in human submandibular gland cells. J Biol Chem 1998; 273: Liu X, Ambudkar I. Characteristics of a Store-operated Calcium-permeable channels, SOCC: sarcoer calcium pump function controls channel gating. J Biol Chem 2001; 276: Barritt GJ. Does a decrease in subplasmalemmal Ca 2+ explain how storeoperated Ca 2+ channels are opened? Cell Calcium 1998; 23:

32 Brereton HM, Harland ML, Froscio M, Petronijevic T, Barritt GJ. Novel variants of voltage-operated calcium channel alpha 1-subunit transcripts in a rat liverderived cell line: deletion in the IVS4 voltage sensing region. Cell Calcium 1997; 22: Brini M, Marsault R, Bastianutto C, Alvarez J, Pozzan T, Rizutto R. Targeted recombinant aequorins: Tools for monitoring [Ca 2+ ] in the various compartments of a living cell. Microsc Res Tech 1999; 46: Burtelow MA, Kaufmann SH, Karnitz LM. Retention of the human Rad9 checkpoint complex in extraction-resistant nuclear complexes after DNA damage. J Biol Chem 2000; 275: Brini M, Marsault R, Bastianutto C, Alvarez J, Pozzan T, Rizutto R. Transfected aequorin in the measurement of cytosolic Ca 2+ concentration ([Ca 2+ ] c ). J Biol Chem 1995; 270: Barrero MJ, Montero M, Alvarez J. Dynamics of [Ca 2+ ] in the ER and cytoplasm of intact Hela cells. A comparative study. J Biol Chem 1997; 272: Shimomura O, Musicki B, Inouye S. Light-emitting properties of recombinant semi-synthetic aequorins and recombinant fluorescein-conjugated aequorin for measuring cellular calcium. Cell Calcium 1993; 14: Webb SE, Lee KW, Karplus E, Miller AL. Localized calcium transients accompany furrow positioning, propagation, and deepening during the early cleavage period of zebrafish embryos. Dev Biol 1997; 192:

33 Chen J, Barritt GJ. Evidence that TRPC1 forms a Ca 2+ permeable channel linked to the regulation of cell volume in liver cells obtained using sirna targeted against TRPC1. Biochem J. 2003; 373: Skehel PA, Fabian-Fine R, Kandel ER. Mouse Vap33 is associated with the ER and microtubules. Proc Natl Acad Sci USA 2000; 97: Montero M, Brini M, Marsault R, Alvarez J, Sitia R, Pozzan T, Rizutto R. Monitoring dynamic changes in free Ca 2+ concentration in the ER of intact cells. EMBO J 1995; 14: Gregory RB, Rychkov G, Barritt GJ. Evidence that 2-aminoethyl diphenylborate is a novel inhibitor of store-operated Ca 2+ channels in liver cells, and acts through a mechanism which does not involve inositol trisphosphate receptors. Biochem J 2001; 354: Bootman MD, Collins TJ, Mackenzie L, Roderick HL, Berridge MJ, Peppiatt CM. 2-aminoethoxydiphenyl borate (2-APB) is a reliable blocker of storeoperated Ca 2+ entry but an inconsistent inhibitor of InsP 3 -induced Ca 2+ release. FASEB J 2002; 16: Brereton HM, Chen J, Rychkov G, Harland ML, Barritt GJ. Maitotoxin activates an endogenous non-selective cation channel and is an effective initiator of the activation of the heterologously expressed htrpc-1 (transient receptor potential) non-selective cation channel in H4-IIE liver cells. Biochim Biophys Acta 2001; 1540:

34 Horstman DA, Tennes KA, Putney JW Jr. ATP-induced calcium mobilization and inositol 1,4,5-trisphosphate formation in H-35 hepatoma cells. FEBS Lett 1986; 204: Corda S, Spurgeon HA, Lakatta EG, Maurizio CC, Ziegelstein RC. ER Ca 2+ depletion unmasks a caffeine-induced Ca 2+ influx in human aortic endothelial cells. Circ Res 1995; 77: Robb-Gaspers LD, Burnett P, Rutter GA, Denton RM, Rizzuto R, Thomas AP. Integrating cytosolic calcium signals into mitochondrial metabolic responses. EMBO J 1998; 17: Fernando KC, Gregory RB, Barritt GJ. Protein kinase A regulates the disposition of Ca 2+ which enters the cytoplasmic space through store-activated Ca 2+ channels in rat hepatocytes by diverting inflowing Ca 2+ to mitochondria. Biochem J 1998; 330: Hofer AM, Fasolato C, Pozzan T. Capacitative Ca 2+ entry is closely linked to the filling state of internal Ca 2+ stores: A study using simultaneous measurements of I crac and intraluminal [Ca 2+ ]. J Cell Biol 1998; 140: Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca 2+ indicators with greatly improved fluorescence properties. J Biol Chem 1985; 260: Blinks JR, Prendergast FG, Allen DG. Photoproteins as biological calcium indicators. Pharm Reviews 1976; 28: 1-93.

35 Kwan C-Y, Putney Jr JW. Uptake and intracellular sequestration of divalent cations in resting and methacholine-stimulated mouse lacrimal acinar cells. J Biol Chem 1990; 265: Wang X, Reznick S, Li P, Liang W, van Breemen C. Ca 2+ removal mechanisms in freshly isolated rabbit aortic endothelial cells. Cell Calcium 2002; 31: Brini M, Bano D, Manni S, Rizutto R, Carafoli E. Effects of PMCA and SERCA pump overexpression on the kinetics of cell Ca 2+ signalling. EMBO J 2000; 19: Braiman A, Priel Z. Intracellular stores maintain stable cytosolic Ca 2+ gradients in epithelial cells by active Ca 2+ redistribution. Cell Calcium 2001; 30: Osada S, Okano Y, Saji S, Nozawa Y. Spontaneous Ca 2+ release from a caffeine and ryanodine-sensitive intracellular Ca 2+ store in freshly prepared hepatocytes. Hepatology 1994; 19: Lily LB, Gollan JL. Ryanodine-induced calcium release from hepatic microsomes and permeabilized hepatocytes. Am J Physiol 1995; 268: Komazaki S, Ikemoto T, Takeshima H, Iino M, Endo M, Nakamura H. Morphological abnormalities of adrenal gland and hypertrophy of liver in mutant mice lacking ryanodine receptors. Cell Tissue Res 1998; 294:

36 McNulty TJ, Taylor CW. Caffeine-stimulated Ca 2+ release from the intracellular stores of hepatocytes is not mediated by ryanodine receptors. Biochem J 1993; 291: Fernando KC, Barritt GJ. Characterisation of the inhibition of the hepatocyte receptor-activated Ca 2+ inflow system by gadolinium and SK & F Biochim Biophys Acta, 1994; 1222: Bolsover S, Silver RA. Artifacts in calcium measurement: recognition and remedies. Trends Cell Biol 1991; 1: Neher E. Usefulness and limitations of linear approximations to the understanding of Ca ++ signals. Cell Calcium 1998; 24: Alonso MT, Chamero P, Villalobos C, Garcia-Sancho J. Fura-2 antagonises calcium-induced calcium release. Cell Calcium 2003; 33: Rembold CM, Van Riper DA, Chen XL. Focal [Ca 2+ ] i increases detected by aequorin but not by fura-2 in histamine- and caffeine-stimulated swine carotid artery. J Physiol 1995; 488: Rembold CM, Chen XL. The buffer barrier hypothesis, [Ca 2+ ] i homogeneity, and sarcoplasmic reticulum function in swine carotid artery. J Physiol 1998; 513.2: Alonso MT, Montero M, Carnicero E, Garcia-Sancho J, Alvarez J. Subcellular Ca 2+ dynamics measured with targeted aequorin in chromaffin cells. Ann New York Acad Sci 2002; 971:

37 Woods NM, Dixon CJ, Cuthbertson KSR, Cobbold PH. Modulation of free Ca oscillations in single hepatocytes by changes in extracellular K +, Na + and Ca 2+. Cell Calcium 1990; 11: Parekh AB, Fleig A, Penner R. The store-operated calcium current I CRAC : nonlinear activation by InsP 3 and dissociation from calcium release. Cell 1997; 89: Hartmann J, Verkhratsky A. Relations between intracellular Ca 2+ stores and store-operated Ca 2+ entry in primary cultured human glioblastoma cells. J Physiol 1998; 513.2: Turner H, Fleig A, Stokes A, Kinet J-P, Penner R. Discrimination of intracellular calcium store subcompartments using TRPV1 (transient receptor potential channel, vanilloid subfamily member 1) release channel activity. Biochem J 2003; 371: Mitchell KJ, Pinton P, Varadi A, Tacchetti C, Ainscow EK, Pozzan T, Rizzuto R, Rutter GA. Dense core secretory vesicles revealed as a dynamic Ca 2+ store in neuroendocrine cells with a vesicle-associated membrane protein aequorin chimera. J Cell Biol 2001; 155:

38 -38- CAPTIONS TO FIGURES Fig. 1 A schematic representation of the proposed relationship between storeoperated Ca 2+ channels (SOCs) in the plasma membrane, the endoplasmic reticulum, and the SERCAs. The locations of aequorin targeted to the cytoplasmic space (cytaeq) and ER (eraeq) are indicated. Also shown are two proposed sub-regions of the ER, the small ER Ca 2+ stores (proposed to be located in close proximity to the SOCs in the plasma membrane and to contain thapsigargin-insensitive SERCAs) and the bulk ER Ca 2+ stores (containing thapsigargin-sensitive SERCAs). It is proposed that Ca 2+ diffuses slowly between these two types of ER Ca 2+ store, and that the activation of SOCs requires a decrease in [Ca 2+ ] er in the small ER Ca 2+ stores. The space between the SOCs and the small ER Ca 2+ stores is designated the subplasmalemmal space, and the majority of the cytoplasmic space the deep cytoplasmic space. The broken lines and arrows indicate that Ca 2+ that enters the subplasmalemmal space can diffuse into the deep cytoplasmic space or be transported into the ER and mitochondria. The pathway will depend on the activity of the SERCAs, the local concentrations of Ca 2+ and the location of the mitochondria. Previous studies by others upon which the scheme is based are summarised in references [14] and [51-53]. Fig. 2 Intracellular locations of cytaeq (A, B) and eraeq (C-F) in H4-IIE cells stably-transfected with cdna encoding cytaeq and eraeq, respectively. (A) cytaeq cells labelled by immunocytochemistry with the monoclonal anti-ha1 antibody. (B) cytaeq cells treated as in A but with the omission of the anti-ha1 antibody. (C-E) The same eraeq cell double-stained with monoclonal anti-ha1

39 -39- antibody (C) and with polyclonal anti-calnexin antibody (D). (E) The merged images of (C) and (D). (F) An eraeq cell treated as in (C) but with omission of the anti-ha1 and anti-calnexin antibodies and viewed using the FITC filter. The scale bar represents 5 µm. The results are representative of those obtained from 5 cells which each gave results similar to those shown. Fig. 3 Transient increases in [Ca 2+ ] cyt induced by the addition of DBHQ to cytaeq cells. Holo-aequorin was reconstituted by incubation of cytaeq cells with coelenterazine in the presence of 1 mm Ca 2+ o, as described in the Materials and Methods. After washing, the cells were incubated in KRB either in the absence of added Ca 2+ o or in the presence of 2 mm Ca 2+ o, as indicated on the figures. For experiments involving Gd 3+, Na 2 PO 4 was omitted from the KRB for incubations in both the presence and absence of Gd 3+. (A) DBHQ (20 µm) addition in the presence (broken line) or absence (solid line) of 2 mm Ca 2+ o. (B) DBHQ (20 µm) addition in the presence of 2 mm Ca 2+ o and in the presence (broken line) or absence (solid line) of 10 µm Gd 3+. (C) DBHQ (20 µm) addition in the absence of added Ca 2+ o and in the presence (broken line) or absence (solid line) of 10 µm Gd 3+. The traces shown are the means ± SEM of 3-8 experiments. For the results shown in A and B, comparison of the values of [Ca 2+ ] cyt at each of 105, 165 and 225 sec after DBHQ addition with the value of [Ca 2+ ] cyt just before DBHQ addition (Student s t-test for unpaired samples) showed no significant difference. Fig. 4 Changes in [Ca 2+ ] cyt and [Ca 2+ ] er induced by the addition of Ca 2+ o to cytaeq and eraeq cells previously incubated in the absence of added Ca 2+ o. cytaeq or eraeq cells were treated with DBHQ (except in (C)), and holo-aeq was reconstituted

40 -40- by incubation with coelenterazine (cytaeq) or coelenterazine n (eraeq) in the presence of 1 mm EGTA, as described in the Materials and Methods. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+. For experiments involving Gd 3+, EGTA and Na 2 PO 4 were both omitted from the KRB for the experiments conducted in the presence and absence of Gd 3+. (A) The addition of 2 mm Ca 2+ o to eraeq cells and the effect of Gd 3+ (10 µm; present at the beginning of the incubation) on the increase in [Ca 2+ ] er induced by addition of 2 mm Ca 2+ o. (B) The addition of 2 mm Ca 2+ o to cytaeq cells. (C) The sequential addition of 1 and 10 mm Ca 2+ o to cytaeq cells (not pre-treated with DBHQ). (D) The effect of Gd 3+ (10 µm; present at the beginning of the incubation) on the transient increase in [Ca 2+ ] cyt induced by addition of 2 mm Ca 2+ o to cytaeq cells. The traces shown are the means ± SEM of 5-11 experiments. The degrees of significance for comparison (Student s t-test for unpaired samples) of the values of [Ca 2+ ] cyt (with and without Gd 3+ ) at the time shown (*) in (A) and at the peak (*) in (D) are P and P 0.01, respectively. In B, comparison of the values of [Ca 2+ ] cyt at 15 sec after Ca 2+ o addition with the value of [Ca 2+ ] cyt immediately before Ca 2+ o addition gave P In C, comparison of values of [Ca 2+ ] cyt after Ca 2+ addition with the value of [Ca 2+ ] cyt immediately before the first Ca 2+ o addition gave P and P 0.05 at 15 and 30 sec, respectively, after the first Ca 2+ o addition, and P at 30 sec after the second Ca 2+ addition. In D, for either Gd 3+ present or Gd 3+ absent, there was no significant difference between the value of [Ca 2+ ] cyt at 15, 30, 45, 60 or 75 sec after Ca 2+ o addition and the value of [Ca 2+ ] cyt immediately before Ca 2+ o addition. Fig. 5 Images of eraeq-generated luminescence from a single H4-IIE cell stably expressing eraeq following the addition of extracellular Ca 2+. Luminescence images

41 -41- (A-E) and corresponding bright-field images (A -E ) of a single cell before (A), immediately after (B), and 10 (C), 20 (D) and 30 (E) sec after the addition of 2 mm Ca 2+ o. eraeq cells were treated with DBHQ, and holo-aeq was reconstituted by incubation with coelenterazine f in the presence of 1 mm EGTA, as described in Materials and Methods. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+. Each image represents 60 sec of accumulated luminescence with a 10 sec step between each image as indicated in B, C, D and E. The colour scale indicates luminescent flux in photons per pixel and the scale bar represents 10 µm. The results shown are representative of those obtained from 5 cells which gave similar images. Fig. 6 The effects of ATP (A, B), caffeine (C) and ionomycin (D) on the [Ca 2+ ] cyt spike and increase in [Ca 2+ ] er induced by addition of Ca 2+ o. CytAEQ or eraeq cells were treated with DBHQ and holo-aeq was reconstituted by incubation with coelenterazine (cytaeq) or coelenterazine n (eraeq) in the presence of 1 mm EGTA, as described in the Materials and Methods. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+ o. (A, B) The addition of 2 mm Ca 2+ o to cytaeq (A) or eraeq (B) cells in the presence and absence of ATP (100 µm; added at the beginning of the incubation). The inset in (A) shows the trace at the addition of ATP. (C) The addition of 2 mm Ca 2+ o to eraeq cells in the presence and absence of caffeine (1 mm; added at the beginning of the incubation). (D) The addition of 2 mm Ca 2+ o to cytaeq cells in the presence and absence of ionomycin (10 µm; added at the beginning of the incubation). The inset in (D) shows the region at the addition of ionomycin. The traces shown are the means ± SEM of 3-11

42 -42- experiments. The degree of significance for comparison of the values of [Ca 2+ ] cyt at the times shown (*) are P Fig. 7 The effects of loading cells with BAPTA (A, B) or EGTA (C) on the [Ca 2+ ] cyt spike (A, C) and increase in [Ca 2+ ] er (B) induced by the addition of Ca 2+ o. cytaeq or eraeq cells were treated with DBHQ and holo-aeq reconstituted by incubation with coelenterazine (cytaeq) or coelenterazine n (eraeq) in the presence of 1 mm EGTA, as described in the Materials and Methods. Cells were loaded with dimethyl BAPTA (A), dibromo BAPTA (B) or EGTA (C) by incubation with 1 mm of the acetoxymethylester of dimethyl BAPTA (A), dibromo BAPTA (B) or EGTA (C) for 60 min at 22 C after holo-aeq reconstitution. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+. Ca 2+ o (2 mm) was present during the period indicated. The data are expressed as means ± SEM of 5-12 experiments. The degrees of significance for comparison of the values of [Ca 2+ ] cyt at the times shown (*) are P 0.01 (A) and P (B and C). Fig. 8 The effects of thapsigargin and DBHQ (present during the Ca 2+ o addition) on the [Ca 2+ ] cyt spike and increase in [Ca 2+ ] er induced by addition of Ca 2+ o. CytAEQ or eraeq cells were treated with DBHQ and holo-aeq was reconstituted by incubation with coelenterazine (cytaeq) or coelenterazine n (eraeq) in the presence of 1 mm EGTA, as described in the Materials and Methods. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+. (A, B) cytaeq (A) or eraeq (B) cells incubated in the presence or absence of thapsigargin (10 µm; added at the beginning of the incubation). (C, D) cytaeq (C) or eraeq (D) cells were incubated in the presence or absence of DBHQ (10 µm) (added at the beginning of the

43 -43- incubation). Ca 2+ (2 mm) was added as shown in the figures. The data are expressed as means ± SEM of 4-13 experiments. In B, the degrees of significance (Student s t- test for unpaired samples) for comparison of the values of [Ca 2+ ] er (presence or absence of thapsigargin) at the time points indicated (*) are P In A and C, for results in either the presence or absence of thapsigargin or DBHQ, the degrees of significance for comparison of the value of [Ca 2+ ] cyt at 15, 30 or 45 sec after Ca 2+ o addition with the value of [Ca 2+ ] cyt immediately before Ca 2+ o addition were each P 0.05 to P In B (for results in the presence of thapsigargin) and D (for results in the presence of DBHQ), the degrees of significance for comparison of the values of [Ca 2+ ] er at 1, 2, or 3 min after Ca 2+ o addition with the value of [Ca 2+ ] er immediately before Ca 2+ o addition were each P 0.05 to P Fig. 9 The effect of thapsigargin on the increase in [Ca 2+ ] cyt induced by CCCP added after Ca 2+ o addition to cytaeq cells (A), and the effect of CCCP, added before Ca 2+ o, on the Ca 2+ o-induced spike of [Ca 2+ ] cyt in the presence and absence of thapsigargin (B). CytAEQ cells were treated with DBHQ and holo-aeq reconstituted by incubation with coelenterazine in the presence of 1 mm EGTA, as described in the Materials and Methods. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+. (A) CCCP (10 µm) was added after 2 mm Ca 2+ o. The inset shows the increase in [Ca 2+ ] cyt following the addition of Ca 2+ o. (B) CCCP (10 µm) was added before 2 mm Ca 2+ o. The inset shows the effects of the addition of CCCP. The results are expressed as the means ± SEM of 3-7 experiments. Fig. 10 Effect of TPEN on Ca 2+ o-induced increases in [Ca 2+ ] cyt and [Ca 2+ ] er. (A) Holo cytaeq was reconstituted by incubation of cytaeq cells with coelenterazine in

44 -44- the presence of 1 mm Ca 2+ o, as described in the Materials and Methods. The cells were initially incubated in KRB in the absence of added Ca 2+. TPEN (400 µm) was added to cytaeq cells before the addition of Ca 2+ o. (B) eraeq cells were treated with DBHQ and holo-aeq reconstituted by incubation with coelenterazine n in the presence of 1 mm EGTA, as described in the Materials and Methods. Cells were initially incubated in KRB which contained 100 µm EGTA and no added Ca 2+. TPEN (400 µm) was added before the addition of 2 mm Ca 2+ o. (C) In cytaeq cells not subjected to DBHQ pre-treatment, holo cytaeq was reconstituted by incubation with coelenterazine in the presence of 1 mm EGTA. After washing, the cells were incubated in KRB in the presence of 2 mm Ca 2+ o. TPEN (400 µm) was added after the addition of 2 mm Ca 2+ o. The results are expressed as the means ± SEM of 4-11 experiments. The degrees of significance for comparison of the value of [Ca 2+ ] cyt in (B) at the time indicated (*) is P Fig. 11 Changes in [Ca 2+ ] cyt measured using fura-2 under conditions similar to those employed using aequorin as a sensor for [Ca 2+ ] cyt. (A) DBHQ was added in the presence of Ca 2+ o, and in the presence and absence of Gd 3+ (10 µm). (B) DBHQ was added in the absence of Ca 2+ o, cells were washed to remove DBHQ, then Ca 2+ was added. (C) DBHQ was added in the absence of Ca 2+ o followed by addition of Ca 2+ (while DBHQ was still present). The additions of DBHQ (20 µm) and Ca 2+ o (2 mm) are shown by the horizontal bars. Cells were loaded with fura-2 and fluorescence ratios measured as described in the Materials and Methods. Each trace represents one of 2-4 traces which each gave similar results.

45 -45-

46 -46-

47 -47-

48 -48-

49 -49-