The role of cyclic ADP-ribose in the amplification of subcellular and global Ca2+ signaling upon stimulation of P2Y purinergic receptors was studied in 3T3 fibroblasts. Either (1) 3T3 fibroblasts (CD38− cells), (2) 3T3 fibroblasts preloaded by incubation with extracellular cyclic ADP-ribose (cADPR), (3) 3T3 fibroblasts microinjected with ryanodine, or (4) 3T3 fibroblasts transfected to express the ADP-ribosyl cyclase CD38 (CD38+ cells) were used. Both preincubation with cADPR and CD38 expression resulted in comparable intracellular amounts of cyclic ADP-ribose (42.3 ± 5.2 and 50.5 ± 8.0 pmol/mg protein). P2Y receptor stimulation of CD38− cells yielded a small increase of intracellular Ca2+ concentration and a much higher Ca2+ signal in CD38-transfected cells, in cADPR-preloaded cells, or in cells microinjected with ryanodine. Confocal Ca2+ imaging revealed that stimulation of ryanodine receptors by cADPR or ryanodine amplified localized pacemaker Ca2+ signals with properties resembling Ca2+ quarks and triggered the propagation of such localized signals from the plasma membrane toward the internal environment, thereby initiating a global Ca2+ wave.
Cyclic ADP-ribose (cADPR), a potent Ca2+ mobilizer from ryanodine-sensitive calcium stores and functionally active in a wide variety of cell types, is generated from NAD+ as substrate by a family of multifunctional enzymes designated ADP-ribosyl cyclases (Guse, 2002; Lee, 2001, 2002). Two ecto-ADP-ribosyl cyclases have been cloned and characterized: the transmembrane type II glycoprotein CD38 and the GPI-anchored protein CD157 (BST-1). Cytosolic ADP-ribosyl cyclase activities have been observed in the marine mollusk Aplysia californica, where the first cyclase was cloned and fully characterized (Lee, 2002), but also in sea urchin eggs (Graeff et al., 1998), human T-lymphocytes (Guse et al., 1999), human blood mononuclear cells (Bruzzone et al., 2003), rat pancreatic acinar cells (Sternfeld et al., 2003), and bovine brain (Matsumura and Tanuma, 1998).
cADPR is but one of various signal metabolites (cADPR, d-myo-inositol 1,4,5-trisphosphate [IP3], and nicotinic acid adenine dinucleotide phosphate [NAADP]) that can release Ca2+ from specific internal stores, in some cases coproduced in the same cell type (for review see da Silva and Guse, 2000). Many cell types have been reported to harbour IP3-, cADPR-, and NAADP-sensitive stores (Albrieux et al., 1998; Guse et al., 1999; Berg et al., 2000; Cancela et al., 2000; Santella et al., 2000; Churchill and Galione, 2001; Hoesch et al., 2002; Brailoiu et al., 2003). The complex spatio-temporal patterns of functional interplay among these Ca2+-mobilizing second messengers and their target receptors represent a central issue in order to elucidate the mechanisms that underlie mobilization of Ca2+ from the different stores, thereby affecting fundamental and diverse cell functions (Meldolesi and Pozzan, 1998; Berridge et al., 2000; Meldolesi, 2002; Carafoli, 2003).
A long recognized paradox of the NAD+/cADPR system is its compartmentation in several mammalian cell types (for review see De Flora et al., 2002). Thus, for instance, the exposure of intact cells to extracellular cADPR has been shown to upgrade the functional response to different agonists (De Flora et al., 1996; Podestà et al., 2000; Franco et al., 2001a; Zocchi et al., 2001). Studies performed on murine 3T3 fibroblasts revealed that internalization of extracellular cADPR occurs through a number of equilibrative and concentrative nucleoside transporters (Guida et al., 2002) and that influx of cADPR into intact 3T3 cells is paralleled by a sustained increase of the basal intracellular Ca2+ concentration ([Ca2+]i) (Franco et al., 2001b). Indeed, comparable increases of the basal [Ca2+]i are also observed after “de novo” expression of CD38 in 3T3 cells, as a result of the related generation of intracellular cADPR, which is responsible for the doubling of [Ca2+]i in CD38+ 3T3 compared with antisense-transfected (CD38−) or wild-type cells (Zocchi et al., 1998).
Murine 3T3 fibroblasts seem to represent a good experimental system to study the interplay between the Ca2+-mobilizing metabolites cADPR and IP3 for a number of reasons: (1) 3T3 cells respond to extracellular micromolar ATP with an IP3-dependent calcium release mediated by P2Y purinergic receptors (Giovannardi et al., 1992); (2) cADPR can be internalized by intact 3T3 fibroblasts across the above-mentioned nucleoside transporters, without the need to permeabilize the cells (Guida et al., 2002); (3) sense- and antisense CD38–transfected cells have a significantly different [Ca2+]i, due to presence or absence, respectively, of intracellular cADPR (Zocchi et al., 1998); and (4) both IP3- and cADPR-sensitive calcium stores are present in this cell type (Giovannardi et al., 1992; Zocchi et al., 1998).
Therefore, we investigated whether the presence or absence of intracellular cADPR can trigger distinctive Ca2+ responses to ATP stimulation in 3T3 fibroblasts. The results obtained indicate that cADPR and IP3 act in a functionally and spatially coordinated fashion and specifically that the presence of intracellular cADPR elicits a clearcut amplification of IP3-mediated [Ca2+]i responses to extracellular ATP.
Different [Ca2+]i responses to ATP in CD38+ and CD38− 3T3 fibroblasts
Intact sense (CD38+)- and antisense (CD38−)-transfected 3T3 fibroblasts were comparatively challenged with 100 μM ATP, a concentration known to stimulate P2Y receptors (Giovannardi et al., 1992; Di Virgilio et al., 2001). The immediate increase of cytosolic [Ca2+]i was remarkably different in the two cell types, with the CD38+ cells exhibiting much higher peak and plateau responses to ATP (Fig. 1 A). When ATP was supplemented in the presence of EGTA, the extent of [Ca2+]i increase was almost superimposable to that recorded in a Ca2+-containing buffer, thus indicating release from intracellular stores as the main underlying mechanism (Fig. 1 B). On the contrary, when CD38− and CD38+ cells were stimulated with 3 mM ATP (a concentration that triggers the P2X receptors, see Di Virgilio et al., 2001), the two cell populations showed quite comparable [Ca2+]i increases. These were abolished by the presence of EGTA in the buffer, therefore demonstrating that calcium influx follows stimulation of the P2X receptors (not depicted). Thus, the [Ca2+]i increases elicited by calcium influx across ATP-gated ion channels (P2X receptors) are not influenced by CD38 expression in 3T3 fibroblasts.
Causal role of intracellular cADPR in the different [Ca2+]i responses to ATP of CD38+ and CD38− 3T3 fibroblasts
A distinctive feature between the CD38− and CD38+ cells is the presence of intracellular cADPR in the latter cell population, as a consequence of the expression of ADP-ribosyl cyclase activity (Zocchi et al., 1998). The amount of intracellular cADPR in CD38+ 3T3 fibroblasts was assayed with a highly sensitive procedure of enzymatic cycling (Graeff and Lee, 2002) and estimated to be 50.48 ± 8.03 pmol/mg protein (n = 8). Conversely, the concentration of cADPR in CD38− cells was hardly detectable (0.25 ± 0.11 pmol/mg protein, n = 9). The corresponding levels of ectocellular ADP-ribosyl cyclase activity, taken as a measure of CD38 content, were 91.25 ± 8.83 (n = 4) and 0.28 ± 0.06 (n = 5) pmol cADPR/min/mg, respectively, in the CD38+ and CD38− cells. To investigate whether the different response to 100 μM ATP could be due to the presence of cADPR, intact CD38− cells were preincubated for 10 min with extracellular cADPR (50 μM), which was recently reported to be internalized by these cells through equilibrative and concentrative nucleoside transporters (Guida et al., 2002).
After preincubation with cADPR, the CD38− 3T3 fibroblasts acquired an ATP-evoked Ca2+ release that was quantitatively comparable to that recorded in the CD38+ cells (Fig. 2 A). The content of intracellular cADPR under these conditions was 42.32 ± 5.24 pmol/mg protein (n = 4).
To establish optimal experimental conditions, the CD38− 3T3 cells were exposed to extracellular cADPR for different time intervals (from 10 min to 18 h). An incubation of 10 min proved to be sufficient to elicit maximal Ca2+ responses to extracellular ATP, and therefore this time was routinely used in subsequent experiments. The observation that a time of minutes was sufficient for these responses to take place indicated that no long-term events downstream of the appearance of intracellular cADPR are causally involved in the Ca2+ response to ATP observed in the cADPR-loaded CD38− cells (Fig. 2 A) or in the CD38-transfected fibroblasts (Fig. 1).
Next, we investigated a possible increase of the intracellular cADPR concentration in CD38+ cells as a consequence of their pulse exposure to ATP. To this purpose, cells were incubated for 0, 10, and 30 s in the presence of 100 μM ATP; the intracellular concentrations of cADPR, however, were not significantly modified after this treatment, i.e., 43.21 ± 2.05, 40.11 ± 4.31, and 45.94 ± 2.45 pmol cADPR/mg protein (n = 3) after 0, 10, and 30 s of treatment, respectively.
To check the specificity of the effect of cADPR, CD38+ 3T3 cells were preincubated for 2 h in the presence of 100 μM 8-Br-cADPR, a membrane-permeant cADPR antagonist (Walseth and Lee, 1993). Under these conditions, the increase of the [Ca2+]i observed after stimulation with ATP was markedly inhibited (Fig. 2 B).
It is well documented that 100 μM ATP evokes a calcium release from IP3-sensitive stores (Giovannardi et al., 1992; Di Virgilio et al., 2001). Therefore, CD38+ 3T3 cells were first incubated either with 2-APB, a membrane-permeant IP3 antagonist, or with U73212, a membrane-permeant inhibitor of PLC. As shown in Fig. 3 A, 2-APB (250 μM) completely abrogated the response to 100 μM ATP, whereas the response to 3 mM ATP was not impaired (not depicted). Thus, the use of 2-APB was instrumental for discriminating between the two different mechanisms that underlie the [Ca2+]i increases after stimulation of the P2Y receptors (Ca2+ release) and of the P2X receptors (Ca2+ influx), respectively. Likewise, U73122 at 1 μM completely inhibited the Ca2+ release triggered by 100 μM ATP, while the same concentration of the inactive analogue U73343 proved to be ineffective (Fig. 3 B).
In an attempt to elucidate the mechanisms responsible for the difference in the ATP-stimulated global Ca2+ signals between wild-type 3T3 cells and cells with increased cADPR contents (obtained either by extracellular addition of cADPR or by transfection with CD38), rapid confocal Ca2+ imaging experiments were performed (Kunerth et al., 2003). Besides confirming the previous results, single cell Ca2+ imaging clearly defined that the limited average response induced by extracellular ATP in CD38− 3T3 cells (Fig. 1) was due to a full response occurring in very few cells (2/23), with the same amplitude observed in CD38+ cells. In contrast, most of the CD38+ cells were responsive to ATP (17/20). This agrees with the earlier observation that in each individual wild-type 3T3 cell, the ATP-induced Ca2+ rise occurs in an all-or-none fashion (Giovannardi et al., 1992). When the CD38− cells were preloaded with cADPR by incubation with 50 μM extracellular concentrations of this cyclic nucleotide, ∼60% of cells acquired the ability to respond to ATP (16/27). On the contrary, most CD38+ cells, after preincubation with 100 μM 8-Br-cADPR, lost their responsiveness to ATP, and only in one third of these cells was the release evoked with the same amplitude (7/21).
Cooperation between cADPR and IP3 in the [Ca2+]i response to extracellular ATP
Detailed analysis of high-resolution Ca2+ images acquired under the four different conditions, (1) CD38− cells, (2) cADPR-loaded CD38− cells, (3) CD38− cells microinjected with ryanodine, and (4) CD38+ cells, revealed fundamental differences in the spatio-temporal patterns of Ca2+ signaling (Figs. 4 and 5). While CD38− cells showed a low [Ca2+]i throughout the cell (Fig. 4 A, left), a slightly increased [Ca2+]i was observed in unstimulated CD38− cells preincubated with cADPR (Fig. 4 B, left) or microinjected with ryanodine 15 min before (Fig. 4 C, left), or in CD38+ cells (Fig. 4 D, left). Upon stimulation by ATP, only very small local Ca2+ signals close to the plasma membrane were observed in the CD38− cells (Fig. 4 A, 117.9 s), whereas CD38− cells either preincubated with cADPR or microinjected with ryanodine, or CD38 transfectants, developed a rapid and global response that travelled across the whole cell as a regenerating wave (Fig. 4, B–D). In CD38− cells preincubated with cADPR or microinjected with ryanodine, and in CD38-transfectants, too, Ca2+ waves started at specific hot-spots at the cell border and travelled toward the perinuclear region where a significant amplification occurred (Fig. 4, B–D). Interestingly, ryanodine receptors (RyRs) were localized in high density in the perinuclear region (not depicted), compatible with their involvement in the amplification process. Detailed analyses of [Ca2+]i distribution in differently localized regions of interest (ROIs) in the cell further illustrated the patterns of wave propagation and amplification observed when cADPR was present in the cells, either after direct loading or in the CD38-expressing cells (Fig. 4 E). Microinjection of an activating concentration of ryanodine 15 min before addition of ATP mimicked the effect of both cADPR preloading or transfection of CD38 (Fig. 4, C and E), suggesting that cADPR indeed acts on the RyR.
For the onset of global Ca2+ waves to occur, local subcellular Ca2+ signals are required as pacemaker signals to initiate the global signal (Bootman et al., 1997; Meldolesi, 2002). Analysis of subcellular Ca2+ signals before stimulation of P2Y receptors (basal condition) revealed a small increase in magnitude in CD38− cells previously loaded with cADPR or in CD38+ cells, while microinjection of ryanodine 15 min before had no stimulatory effect (Table I).
After stimulation with ATP, increasing pacemaker signals, localized in proximity of the plasma membrane and a few micrometers inside the cell, were rarely visible in CD38− cells (Fig. 5 A). In contrast, these signals were frequently present in the CD38− cells preloaded with cADPR (Fig. 5 B) or microinjected with ryanodine (Fig. 5 C), or in CD38+ cells (Fig. 5 D). Time course analysis of characteristic pacemaker signals revealed increases in amplitude with time for the latter three conditions (Fig. 5, B–D vs. A), with the slope differing slightly between the three different conditions (Fig. 5, B–D). Quantitative analysis of the pacemaker signals resulted in a very small increase of the amplitude in CD38− cells upon stimulation by ATP (Table I), while in CD38− loaded with cADPR or microinjected with ryanodine 15 min before, significantly higher amplitudes upon ATP stimulation were observed (Table I). The latter values were similar to the ones obtained in CD38+ cells upon P2Y receptor stimulation (Table I). Furthermore, the frequency of the pacemaker signals upon ATP stimulation increased significantly when CD38− cells were loaded with cADPR or microinjected with ryanodine (Table I). Interestingly, a similar frequency was observed in CD38− cells loaded with cADPR and in CD38+ cells, while CD38− cells microinjected with ryanodine 15 min before showed an even more pronounced frequency (Table I). The spatial extension of the pacemaker signals revealed a remarkable similarity under all conditions with signal areas between 0.33 and 0.42 μm2 (Table I), indicating that in the basal phase and in the early pacemaker phase, an increased open probability of RyR, rather than rapid recruitment of further RyR, is the mechanism of signal amplification.
Fibroblasts feature Ca2+ signaling as an important signal transduction system downstream of stimulation of purinergic P2Y receptors (Giovannardi et al., 1992; Zheng et al., 1998; Homolya et al., 1999; Meszaros et al., 2000). However, in the 3T3 wild-type cell line used in our study, the mean ATP-induced Ca2+ signal was low (Fig. 1) and, when analyzed at the single cell level, it occurred only in a minority of cells in an all-or-none-fashion, as previously described (Giovannardi et al., 1992). Similar small Ca2+ signals upon P2Y receptor ligation were also described for the mouse wild-type L-fibroblasts, but they were significantly enhanced by stable overexpression of type 1 IP3 receptor (IP3R) (Davis et al., 1999). These data indicate that a certain threshold of Ca2+ signals generated by the purinoceptor-activated machinery of fibroblasts must be reached to produce global and high magnitude Ca2+ signals, e.g., a global regenerating Ca2+ wave.
In the present study, we pursued an alternative approach to shift the spatio-temporal Ca2+ patterns toward a substantially greater response by triggering the cADPR/Ca2+ signaling pathway. This was obtained in three experimental ways. First, extracellular cADPR was added to CD38− fibroblasts, which express in their plasma membrane both equilibrative and concentrative transporters previously demonstrated to also internalize cADPR (Guida et al., 2002). Second, an activating concentration of ryanodine was microinjected into CD38− fibroblasts 15 to 20 min before stimulation of P2Y receptors. Third, wild-type, constitutively CD38− 3T3 fibroblasts were transfected with human CD38 cDNA (Zocchi et al., 1998). Under these conditions, the CD38− cells acquired the property to respond to stimulation of P2Y receptors like the CD38 transfectants. However, IP3 generation is involved as a necessary initial event but is insufficient to trigger global Ca2+ signaling.
These findings confirm that cADPR, no matter whether being internalized from extracellular sites (Guida et al., 2002), or being generated at the outer surface of the same cell and then actively internalized by transmembrane CD38 (Franco et al., 1998), or being produced in intracellular vesicles and then extruded therefrom to the cytosol (Bruzzone et al., 2001), is in all cases able to access its target Ca2+ stores. These mechanisms have been shown to functionally circumvent topological constraints related to compartmentalization of both NAD+ and cADPR (for review see De Flora et al., 2002).
Furthermore, 3T3 murine fibroblasts proved to represent a profitable model to investigate functional interactions between IP3R and RyR, i.e., a cell type featuring “channel cross-talk” (Patel et al., 2001; Morgan and Galione, 2002). Enhanced activity of either of the two Ca2+ release systems in fibroblasts by overexpression of type 1 IP3R (Davis et al., 1999) or by de novo inducing intracellular cADPR production (Zocchi et al., 1998; this study) resulted in an increased mean [Ca2+]i in unstimulated cells. Analysis of the subcellular Ca2+ distribution by confocal Ca2+ imaging revealed the presence of localized Ca2+ signals with increased Ca2+ concentrations both in L-fibroblasts overexpressing the type 1 IP3R (Davis et al., 1999) and in the CD38− 3T3 cells either preincubated with cADPR or microinjected with ryanodine (Fig. 5; Table I).
Analysis of the subcellular Ca2+ release events in 3T3 fibroblasts revealed amplitudes of 44–98 nM and areas of 0.33–0.42 μm2 (corresponding to diameters of ∼0.57 and 0.65 μm). These values are considerably smaller as compared with typical sparks (amplitude 71–300 nM, diameter 2–5 μm, for an extended list of references see Discussion section of Kunerth et al., 2003). However, so-called “fundamental” Ca2+ signals produced by very few RyRs (possibly one) were described as Ca2+ quarks in skeletal and cardiac muscle (Tsugorka et al., 1995; Lipp and Niggli, 1998). These Ca2+ quarks were characterized by diameters between 0.3 and 0.85 μm and amplitudes of ∼40 nM (Tsugorka et al., 1995; Lipp and Niggli, 1998), values very similar to the ones described here. This indicates that the subcellular Ca2+ release events observed under basal conditions and during the very early pacemaker phase in 3T3 cells are comparable to fundamental Ca2+ quarks that have been observed so far only in excitable cells. Moreover, upon ATP stimulation, the quark-like early pacemaker signals were further increased in amplitude and in frequency, but not in diameter. This indicates that longer and more frequent opening of the RyR channel, but not recruitment of further channels, is the major mechanism for signal amplification by both cADPR and ryanodine in this early pacemaker phase.
In conclusion, it is of remarkable interest that a very similar phenotype, namely a global Ca2+ wave upon P2Y receptor stimulation, could be obtained in fibroblasts either by enhancing the IP3/Ca2+ signaling pathway (Davis et al., 1999) or by providing the cell with the cADPR/Ca2+ signaling system (this study). As overexpression of different IP3R subtypes may increase or decrease dramatically upon certain conditions (Davis et al., 1999), the transition from local to global Ca2+ signals related to increased density of IP3Rs can be of physiological significance.
Also, the cADPR/Ca2+ signaling pathway is susceptible to be widely modulated in RyR-expressing cells. Relevant examples include (1) the massive expression of CD38 that is causally related to retinoic acid–induced granulocytic differentiation of HL60 cells (Munshi et al., 2002); (2) the increased concentrations of intracellular cADPR elicited by lipopolysaccharide in human blood mononuclear cells (Bruzzone et al., 2003); and (3) the delivery of extracellular cADPR by CD38+ neighboring cells across various nucleoside transporters that allow cells negative for CD38, but positive for RyR, to feature cADPR-dependent Ca2+ responses and increased Ca2+-mediated processes (for review see De Flora et al., 2002).
In conclusion, cADPR behaves as a paracrine messenger able to switch different cell types from low to high “excitability” (Franco et al., 2001a,b; Verderio et al., 2001; Zocchi et al., 2001). Our present findings indicate that in 3T3 fibroblasts, the underlying mechanism is the amplification of quark-like subcellular Ca2+ signals by cADPR, both in the basal phase and in the early pacemaker phase.
Materials And Methods
Fura 2-AM and 2-APB were obtained from Calbiochem. All other chemicals were obtained from Sigma-Aldrich.
Determination of intracellular cADPR content
Resting CD38+ and CD38− 3T3 fibroblasts were extracted in 0.5 ml of 0.6 M PCA, and an aliquot of the cell suspension was submitted to protein determination, according to Bradford (1976). In another set of experiments, CD38+ cells were seeded in six-well plates and then extracted by the addition of 0.6 M PCA after 0, 10, and 30 s exposure to 100 μM ATP. Protein determination was performed on cells from wells prepared in parallel. After removal of proteins, cADPR was measured on the neutralized extracts by a highly sensitive enzymatic cycling assay (Graeff and Lee, 2002). The intracellular cADPR concentrations were expressed as pmol/mg protein.
Assay of ADP-ribosyl cyclase activity
ADP-ribosyl cyclase activity was assayed as previously described (Bruzzone et al., 2003). In brief, intact CD38+ and CD38− cells (106) were resuspended in 400 μl of PBS–glucose (10 mM) with 0.1 mM NAD+. At different times (0, 2, 5, 10, and 60 min), 100-μl aliquots were withdrawn, and 220 μl of 0.9 M PCA was added to each aliquot. After deproteinization, PCA was removed, and cADPR content was measured in each aliquot according to the cycling enzymatic assay (Graeff and Lee, 2002). Protein determination was performed on an aliquot of the incubation (Bradford, 1976).
cADPR influx into intact CD38− 3T3 fibroblasts
cADPR influx into intact CD38− cells was performed as previously described (Guida et al., 2002). In brief, cells were harvested and resuspended in 100 μl of Na+ buffer (135 mM NaCl, 6.3 mM K2HPO4, 2.7 mM KCl, 1.5 mM KH2PO4, 0.5 mM MgCl2, 0.9 mM CaCl2, 10 mM glucose, pH 7.4) in the presence of 50 μM cADPR at 22°C for 10 min. The suspension was then centrifuged at 5,000 g for 15 s. Pellets were washed with 1.5 ml of ice-cold appropriate Na+ buffer containing 10 mM uridine (to inhibit loss of internalized cADPR across equilibrative nucleoside transporters, see Guida et al., 2002) and submitted to two consecutive centrifugations as described above to remove the supernatants completely. Pellets were resuspended in 300 μl water, and the samples were sonicated for 30 s at 3 W in ice. Aliquots of 280 μl were deproteinized with 0.6 M perchloric acid (final concentration), and cADPR was detected by the enzymatic cycling assay as described in “Determination of intracellular cADPR content” (Graeff and Lee, 2002). Protein content was determined on 20-μl aliquots according to Bradford (1976).
Calcium measurements in cell populations
Both 3T3+ and 3T3− cells were seeded in 96-well plates (50 × 103 cells/well). Cells were loaded with 10 μM FURA-2/AM (or with Fluo-3/AM) for 30 min in complete medium. Cells were then washed twice with 200 μl of calcium buffer (135 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 5 mM Hepes, pH 7.4), and 100 μl of the same buffer was added in each well. Calcium-free buffer was prepared without CaCl2 and with the addition of 2 mM EGTA. Fluorescence was measured every 2.12 s (excitation, 355 nm and 390 nm, alternatively; emission, 520 nm) using a fluorescence plate reader (Fluostar Optima; BMG Labtechnologies GmbH). The ratio of emitted light after excitation at 355 nm/390 nm was calculated and displayed as a function of time. In the experiments with cADPR-loaded CD38− cells, the cyclic nucleotide (50 μM) was added to the complete medium during incubation with FURA-2/AM (last 10 min). Cells were washed as described above, and 50 μM cADPR was added to the buffer during the calcium measurements. CD38+ cells were preincubated with 8-Br-cADPR for 2 h in complete medium, and FURA-2AM was added during the last 30 min.
Confocal calcium imaging
The cells were cultured overnight in chamber slides consisting of a plastic chamber and a thin glass coverslip. At the day of measurements, the cells were loaded in these chamber slides with FURA-2/AM (10 μM) for 30 min. After loading, the medium was exchanged against a buffer containing 140 mM NaCl, 5 mM KCl, 1 mM MgSO4, 1 mM CaCl2, 1 mM NaH2PO4, 5.5 mM glucose, and 20 mM Hepes (pH 7.4). The chamber slide was mounted on the stage of a fluorescence microscope (Leica DM IRE2).
Ratiometric Ca2+ imaging was done as described in an earlier report (Kunerth et al. 2003). In brief, we used an Improvision imaging system at 100-fold magnification (Leica objective type HCX APO 100x/1.3 OIL U-V-I; numerical aperture 1.3) built around the Leica microscope at room temperature. Illumination at 340 and 380 nm was performed using a monochromator system (Polychromator IV; TILL Photonics). Images were taken with a gray-scale CCD camera (type C4742–95–12ER; operated in 8-bit mode; Hamamatsu). The optimal relation of spatial and temporal resolution for the ratiometric measurements was obtained using the spatial resolution of 512 × 640 pixel, resulting in a pixel size of 0.129 μm/pixel (at 100-fold magnification). The maximal acquisition rate was ∼160 msec for one ratio using Openlab software (v3.09; Improvision; Atherton et al., 1997). Raw data images were stored on hard disk. To obtain digital confocal images, mathematical deconvolution based on the point-spread function was performed using the no-neighbor algorithm (Openlab software, v1.7.8 and v3.0.9; Improvision; Atherton et al., 1997). The pinhole was set to 70% removal of stray light. After deconvolution of the raw data, confocal ratio images (340/380) were constructed pixel by pixel.
Microinjections were performed as previously described (Guse et al., 1997). An Eppendorf microinjection system (transjector type 5246, micromanipulator type 5171) equipped with Femtotips II as pipettes was used. The system was operated in the semiautomatic mode with the following instrumental settings: injection pressure 40 hPa, compensatory pressure 30 hPa, injection time 0.5 s, and velocity of the pipette 600 μm/s. Ryanodine was diluted to its final concentration (100 μM) in intracellular buffer (20 mM Hepes, 110 mM KCl, 2 mM MgCl2, 5 mM KH2PO4, 10 mM NaCl, pH 7.2) and filtered (0.2 μm) before use. A volume amounting to ∼1% of the cell volume was injected, resulting in an intracellular ryanodine concentration of ∼1 μM. Upon ryanodine injection, the cells displayed increased [Ca2+]i; thus, further stimulation by ATP was performed 15–20 min later, when [Ca2+]i had returned to basal values.
We thank I. Moreschi for technical support.
This study was supported in part by the Deutsche Foschungsgemeinschaft (GU 360/2-4 and 2-5 to A.H. Guse), the Werner-Otto-Foundation (to A.H. Guse), the Wellcome Trust (068065 to A.H. Guse), the Deutsche Akademische Austauschdienst (VIGONI program, grant 314-vigoni-dr to A.H. Guse and A. De Flora), the Associazione Italiana per la Ricerca sul Cancro (to A. De Flora), the MIUR-PRIN 2000 (to A. De Flora), the CNR Target Project on Biotechnology (to E. Zocchi), MIUR-FIRB (RBAU019A3C to A. De Flora), MIUR-FIRB (RBNE01ERXR-003 to A. De Flora), Interuniversity Consortium on Biotechnology (CIB to A. De Flora), and the CARIGE Foundation (to A. De Flora).
S. Bruzzone and S. Kunerth contributed equally to this work.
Abbreviations used in this paper: cADPR, cyclic ADP-ribose; IP3, d-myo-inositol 1,4,5-trisphosphate; IP3R, IP3 receptor; ROI, region of interest; RyR, ryanodine receptor.