Spatial distribution and quantitation of free luniinal

Free [Ca] within organelles of permeabilized BHK-2 1 cells was measured using ratio imaging of compartmentalized mag-fura-2. In BHK-21 cells, this dye monitors free [Ca] in principally one type of ATP-dependent Ca-sequestering organelle in which intrastore Ca was released uniformly and entirely by 100 nM thapsigargin or removal of ATP or Ca from the bath, and was reduced by 85% upon treatment with a supramaxiinal dose of InsP3 (6 JAM). Examination of the spatial distribution of InsP3-sensitive Ca stores showed that InsP3 released Ca throughout all regions of the cell, although we often noted a perinuclear region (which we speculate may correspond to the Golgi apparatus) with reduced responsiveness to InsP3. InsP3-induced changes of intraluminal Mg could not be detected. Cyclic ADPribose, ryanodine, caffeine, mitochondrial inhibitors, and GTP, agents known to influence intraorganellar Ca sequestration in other cell types, were all without effect on the mag-fura-2 ratio. In situ calibration of the mag-fura-2 ratio with Ca ionophores revealed that the average free intraorganellar [Ca] was initially 188 ± 21 JAM in the presence of 170 nMfree Ca and3 mM ATP, and was reduced to 25 ± 5 JAM upon stimulation with 6 JAM InsP3. The ionic dependence of the release and reloading process was also investigated. The presence of either K, Na, or Cl could consistently support both InsP3-induced release and the refilling of stores with Ca, but physiological concentrations of HCO3 were effective in sustaining the response in only 24% of cells examined.-Hofer, A. M., Schlue, W.-R., Curci, S., Machen, T. E. Spatial distribution and quantitation of free luminal [Ca] within the InsP3-sensitive internal store of individual BHK-2 1 cells: ion dependence of InsP3-induced Ca release and reloading. FASEB J. 9, 788-798 (1995) Key Word,t: ratio imaging #{149} thapsigargin #{149} Ca ionophores Virtually all eukaryotic cells possess subcellular stores of compartmentalized Ca that are available for mobilization into the cytoplasm during agonist-stimulated Ca signaling events. Ca stores that can be released by inositol 1,4,5trisphosphate (InsP3)5 are nearly ubiquitous, but other types of releasable Ca poois are known to coexist in the same cell type, for example, stores that are sensitive to ryanodine or caffeine, cyclic ADP ribose, or GTP (see, for example, refs 1, 2 for review). The particular complement of Ca-sequestering pools present depends on the cell type. We have developed a fluorescence technique that allows direct measurement of free [Ca] within the InsP3-releasable pool using the Ca indicator mag-fura-2 (3, 4). This probe becomes compartmentalized in intracellular organelles when loaded as the acetoxymethyl ester derivative, where, by virtue of its relatively high Kd for Ca (53 JAM; refs 5, 6) it reports changes in free [Ca] in this space. Our previous studies were performed on gastric epithelilal cells, which have a relatively complex collection of compartments capable of ATP-dependent Ca accumulation as measured by this probe. In addition to the InsP3-sensitive store, gastric cells were shown to have an unusually robust mitochondrial Ca uptake activity, as well as a Ca-sequestering pool that was determined to be distinct from either the InsP-sensitive pool or the mitochondrial store. A disadvantage of using these cells is that the presence of multiple pools complicates the quantitation of free [Ca] in any one given pool, as described in detail in ref 4. ‘Present address: II. Physiologisches Inst., Universitlit des Saarlandes, D-66421, Homberg/Saar, Germany. 2Present address: Inst. fuurZoologie, Universit#{228}t D#{252}sseldorf, D-40255, D#{252}sseldorf, Germany. 3Present address: Istituto di Fisiologia Generale, Universit#{225} degli Studi di Ban, 70126 Ban, Italy. 4To whom correspondence and reprint requests should be addressed, at: Department of Molecular and Cell Biology, Division of Cell and Developmental Biology, 231 ISA, University of California, Berkeley, CA 94720, USA. 5Abbreviations: InsP3, inositol 1,4,5-trisphosphate; SL-O, streptolysin-O; NMG, N-methyl glucamine. RESEARCH COMMUNICATION QUANTITATION OF FREE LUMINAL IN THE INSP3-SENSITIVE STORE 789 A previous paper concluded from measurements on compartmentalized fura-2 that Ca was released uniformly from the internal stores by InsP3, but quantitation of free [Ca] changes was difficult due to the low Li of fura-2 (7). Moreover, there was no consideration of the issue of the dye becoming compartmentalized in multiple pools, with ensuing complications in interpreting the data. Thus, precise determination of the intraluminal free [Ca] of the internal store has not been possible before, but this value is of potential interest to many investigators, e.g., for understanding the kinetics of the InsP3-induced release and subsequent refilling of stores. Furthermore, many luminal enzymes and resident proteins of organelles bind Ca, and in some cases a regulatory role for Ca binding has been suggested. For instance, in the endoplasmic reticulum BiP, protein disulfide isomerase, reticulocalbin, and reticuloplasmins are luminal Ca binding proteins, the location of which may coincide with the InsP3-sensitive store (8-10). It would be interesting to know whether the binding affinities for Ca lie within a range that might suggest regulation by [Ca] changes occurring during the InsP3-induced release. We report here on a fibroblastic cell type, the BHK-21 cell line, that appears to have essentially one functional type of store capable of accumulating Ca in an ATP-dependent manner as measured by the mag-fura-2 technique. The fact that mag-fura-2 monitors [Ca] in mainly one pool in this cell type greatly simplifies the interpretation of data and permits the estimation of the amount of Ca released during InsP3 stimulation. Furthermore, the flattened morphology of these cells makes it possible to examine the spatial distribution of InsP3-responsive stores using imaging techniques and conventional optics. We also investigated the ionic dependence of InsP3-induced Ca release and reloading in BHK-21 cells. Because Ca pumping by the Ca-ATPase is an electrogenic process, it is expected that refilling the internal store will require the movement of a counterion. Similarly, a counterion would also be expected to accompany Ca during the massive release of Ca from the internal store during agonist stimulation. In fact, it has been established in a number of cell types that a permeant ion must be present in order to elicit a response, but the particular ionic requirements for this process and the pharmacological characteristics of these parallel pathways appear to vary depending on cell type. For example, it has been reported that Cl is required for the refilling of ATP-dependent Ca pools in pancreatic and gastric cells (3, 11). The coupling of Ca fluxes to univalent cations, with particular emphasis on the role of K in the release process, has also been examined in liver, brain, and platelets (12-15). To our knowledge, the participation of HCO3 ions has not been investigated. Here we report that Cl, Na, and K all participate in the release and reloading process after InsP3 stimulation in BHK-21 cells. Physiological concentrations of HCO3 only occasionally were able to support the response. MATERIALS AND METHODS Cell culture, dye loading, and permeabilization BHK-21 cells were grown in Eagle’s MEM with Earle’s BSS containing 10% fetal calf serum and 10% Difco tryptose phosphate broth, and were maintained in a humidified incubator at 37#{176}C in the presence of 5% C02/95% air. Cells were plated onto glass coverslips and used the following day for ratio imaging experiments. For dye loading, cells were incubated with 5 mM mag-fura-2-AM in tissue culture medium at 37#{176}C for 20 mm, after which they were placed for 1 mm into cold permeabilization buffer (4#{176}C) containing reduced streptolysmn-O (SL-O). In some experiments cells were permeabilized at 37#{176}C with 1 mM digitonin as described previously (3, 4). Both permeabilization protocols produced equivalent results. Coverslips were mounted immediately into a metal flow-through perfusion chamber that has been described previously in more detail (16) and perfused continuously with “intracellular buffer” at 37#{176}C. SL-O binds to the plasma membrane at 4#{176}C, but is unable to form the perforating pore complex in the plasma membrane until the temperature is raised to 37#{176}C (17). This procedure permits selective permeabilization of the plasma membrane, allowing cytosolic dye to leak out but leaving compartmentalized mag-fura-2 behind in intact organelles. Ratio imaging experiments After permeabilization with SL-O or digitonin, the cells were then ready for ratio imaging measurements, using a system described previously in more detail (16). The chamber containing the coverslip was fitted onto the heated stage of a Zeiss IM 35 microscope and cells were observed with a 40X objective. Because cells were perfused continuously with intracellular buffer (the turnover time of the perfusion chamber was approximately 2 a) both cytosolic mag-fura-2 and and dye that might leak from subcellular compartments were eliminated. Cells were excited alternately at 350 nm and 385 nm, and video images of the resulting fluorescence emission from each excitation wavelength (collected at 510 nm) were recorded by a silicon-intensified target camera (model 66, Dage-MTI, Michigan City, Ind.). Raw images were forwarded to a Gould FD5000 image processor (Cleveland, Ohio) that produced a backgroundcorrected pseudocolor image of the 350/385 ratio throughout the cell. Images were generally acquired every 10 s. The 350/385 ratio is proportional to the free [Ca] or [Mg] in subeellular compartments resistant to permeabilization. The records shown depict the averaged response of 1-10 individual cells in a single experiment; “n” refers to the number of independent experimental runs. Errors are ± SEM. Solutions and materials Permeabilization buffer contained (in mM): 125 KCI, 25 NaCl, 10 HEPES, pH 7.40, and 1.5 units/ml reduced streptolysin-O (Murex Diagnostics Limited, Dartford, England). Intracellular buffer contained (in mM): 125 KC1, 25 NaCl, 10 HEPES, 3 Na2ATP (unless otherwise stated), 0.1 MgCl2, pH 7.20, with free [Ca] clamped to 170 nM using CaEGTA buffers that were prepared by methods described previously (18). Calibration solutions (as in Fig. 4) had the same composition as the intracellular buffer, except that ATP and EGTA were omitted, and the solutions were supplemented with the indicated amount of CaC12 and the Ca/Mg ionophore 4-Br-A23187. Zero Ca calibration buffer contained 0.8 mM EGTA. The free [Ca] already present in our nominally Ca-free solution was estimated to be approximately 5 J.tM. HCO3._-containing solutions (as in Fig. 5B) were prepared with 25 mM added NaHCO3 (NaCI replacement) and were bubbled with 5% COz/95% air. For experiments using Na and/or K-free solutions (as in Figs. 5C, D), N-methyl glucamine (NMG) chloride or choline chloride were used as replacements, and the tris-salt of ATP was used. For Cl-free solutions (as used in Fig. #{163}4), the corresponding gluconate salt was used. The sucrose-based intracellular solutions used in experiments of the type shown in Fig. 6A contained (in mM): 300 sucrose, 10 HEPES, 3 Tris ATP, 0.1 MgCl2, pH 7.20, and 170 nM free Ca. lso-osmolar replacement of sucrose with the indicated amounts of K gluconate, Na gluconate, or NMG Cl were made for experi-