Failure of calcium microdomain generation and pathological consequences

Normal physiological regulation depends on Ca²⁺ microdomains, because there is a need to spatially separate Ca²⁺ regulation of different cellular processes. It is only possible to generate local Ca²⁺ signals transiently; so, there is an important functional link between Ca²⁺ spiking and microdomains. The pancreatic acinar cell provides a useful cell biological model, because of its clear structural and functional polarization. Although local Ca²⁺ spiking in the apical (granular) microdomain regulates fluid and enzyme secretion, prolonged global elevations of the cytosolic Ca²⁺ concentration are associated with the human disease acute pancreatitis, in which proteases in the granular region become inappropriately activated and digest the pancreas and its surroundings. A major cause of pancreatitis is alcohol abuse and it has now been established that fatty acid ethyl esters and fatty acids, non-oxidative alcohol metabolites, are principally responsible for causing the acinar cell damage. The fatty acid ethyl esters release Ca²⁺ from the endoplasmic reticulum and the fatty acids inhibit markedly mitochondrial ATP generation, which prevents the acinar cell from disposing of the excess Ca²⁺ in the cytosol. Because of the abolition of ATP-dependent Ca²⁺ pump activity, all intracellular Ca²⁺ concentration gradients disappear and the most important part of the normal regulatory machinery is thereby destroyed. The end stage is necrosis.     

Aberrant Ca(2+) signalling through acidic calcium stores in pancreatic acinar cells

Pancreatic acinar cells possess a very large Ca²⁺ store in the endoplasmic reticulum, but also have extensive acidic Ca²⁺ stores. Whereas the endoplasmic reticulum is principally located in the baso-lateral part of the cells, although with extensions into the granular area, the acidic stores are exclusively present in the apical part. The two types of stores can be differentiated pharmacologically because the endoplasmic reticulum accumulates Ca²⁺ via SERCA pumps, whereas the acidic pools require functional vacuolar H⁺ pumps in order to maintain a high intra-organellar Ca²⁺ concentration. The human disease acute pancreatitis is initiated by trypsinogen activation in the apical pole and this is mostly due to either complications arising from gall bladder stones or excessive alcohol consumption. Attention has therefore been focussed on assessing the acute effects of bile acids as well as alcohol metabolites. The evidence accumulated so far indicates that bile acids and fatty acid ethyl esters - the non-oxidative products of alcohol and fatty acids - exert their pathological effects primarily by excessive Ca²⁺ release from the acidic stores. This occurs by opening of the very same release channels that are also responsible for normal stimulus-secretion coupling, namely inositol trisphosphate and ryanodine receptors. The inositol trisphosphate receptors are of particular importance and the results of gene deletion experiments indicate that the fatty acid ethyl esters mainly utilize sub-types 2 and 3.     

Two different but converging messenger pathways to intracellular Ca(2+) release: the roles of nicotinic acid adenine dinucleotide phosphate, cyclic ADP-ribose and inositol trisphosphate

Hormones and neurotransmitters mobilize Ca(2+) from the endoplasmic reticulum via inositol trisphosphate (IP(3)) receptors, but how a single target cell encodes different extracellular signals to generate specific cytosolic Ca(2+) responses is unknown. In pancreatic acinar cells, acetylcholine evokes local Ca(2+) spiking in the apical granular pole, whereas cholecystokinin elicits a mixture of local and global cytosolic Ca(2+) signals. We show that IP(3), cyclic ADP-ribose and nicotinic acid adenine dinucleotide phosphate (NAADP) evoke cytosolic Ca(2+) spiking by activating common oscillator units composed of IP(3) and ryanodine receptors. Acetylcholine activation of these common oscillator units is triggered via IP(3) receptors, whereas cholecystokinin responses are triggered via a different but converging pathway with NAADP and cyclic ADP-ribose receptors. Cholecystokinin potentiates the response to acetylcholine, making it global rather than local, an effect mediated specifically by cyclic ADP-ribose receptors. In the apical pole there is a common early activation site for Ca(2+) release, indicating that the three types of Ca(2+) release channels are clustered together and that the appropriate receptors are selected at the earliest step of signal generation.     

Ryanodine receptors contribute to bile acid-induced pathological calcium signaling and pancreatitis in mice

Biliary pancreatitis is the most common etiology for acute pancreatitis, yet its pathophysiological mechanism remains unclear. Ca(2+) signals generated within the pancreatic acinar cell initiate the early phase of pancreatitis, and bile acids can elicit anomalous acinar cell intracellular Ca(2+) release. We previously demonstrated that Ca(2+) released via the intracellular Ca(2+) channel, the ryanodine receptor (RyR), contributes to the aberrant Ca(2+) signal. In this study, we examined whether RyR inhibition protects against pathological Ca(2+) signals, acinar cell injury, and pancreatitis from bile acid exposure. The bile acid tauro-lithocholic acid-3-sulfate (TLCS) induced intracellular Ca(2+) oscillations at 50 μM and a peak-plateau signal at 500 μM, and only the latter induced acinar cell injury, as determined by lactate dehydrogenase (LDH) leakage. Pretreatment with the RyR inhibitors dantrolene or ryanodine converted the peak-plateau signal to a mostly oscillatory pattern (P < 0.05). They also reduced acinar cell LDH leakage, basolateral blebbing, and propidium iodide uptake (P < 0.05). In vivo, a single dose of dantrolene (5 mg/kg), given either 1 h before or 2 h after intraductal TLCS infusion, reduced the severity of pancreatitis down to the level of the control (P < 0.05). These results suggest that the severity of biliary pancreatitis may be ameliorated by the clinical use of RyR inhibitors.     

Fatty acids, alcohol and fatty acid ethyl esters: Toxic Ca signal generation and pancreatitis

Pancreatitis, a potentially fatal disease in which the pancreas digests itself as well as its surroundings, is a well recognized complication of hyperlipidemia. Fatty acids have toxic effects on pancreatic acinar cells and these are mediated by large sustained elevations of the cytosolic Ca²⁺ concentration. An important component of the effect of fatty acids is due to inhibition of mitochondrial function and subsequent ATP depletion, which reduces the operation of Ca²⁺-activated ATPases in both the endoplasmic reticulum and the plasma membrane. One of the main causes of pancreatitis is alcohol abuse. Whereas the effects of even high alcohol concentrations on isolated pancreatic acinar cells are variable and often small, fatty acid ethyl esters – synthesized by combination of alcohol and fatty acids – consistently evoke major Ca²⁺ release from intracellular stores, subsequently opening Ca²⁺ entry channels in the plasma membrane. The crucial trigger for pancreatic autodigestion is intracellular trypsin activation. Although there is still uncertainty about the exact molecular mechanism by which this Ca²⁺-dependent process occurs, progress has been made in identifying a subcellular compartment – namely acid post-exocytotic endocytic vacuoles – in which this activation takes place.     

Microtubules regulate local Ca2+ spiking in secretory epithelial cells

The role of the cytoskeleton in regulating Ca(2+) release has been explored in epithelial cells. Trains of local Ca(2+) spikes were elicited in pancreatic acinar cells by infusion of inositol trisphosphate through a whole cell patch pipette, and the Ca(2+)-dependent Cl(-) current spikes were recorded. The spikes were only transiently inhibited by cytochalasin B, an agent that acts on microfilaments. In contrast, nocodazole (5-100 micrometer), an agent that disrupts the microtubular network, dose-dependently reduced spike frequency and decreased spike amplitude leading to total blockade of the response. Consistent with an effect of microtubular disruption, colchicine also inhibited spiking but neither Me(2)SO nor beta-lumicolchicine, an inactive analogue of colchicine, had any effect. The microtubule-stabilizing agent, taxol, also inhibited spiking. The nocodazole effects were not due to complete loss of function of the Ca(2+) signaling apparatus, because supramaximal carbachol concentrations were still able to mobilize a Ca(2+) response. Finally, as visualized by 2-photon excitation microscopy of ER-Tracker, nocodazole promoted a loss of the endoplasmic reticulum in the secretory pole region. We conclude that microtubules specifically maintain localized Ca(2+) spikes at least in part because of the local positioning of the endoplasmic reticulum.     

Bile acids induce Ca2+ release from both the endoplasmic reticulum and acidic intracellular calcium stores through activation of inositol trisphosphate receptors and ryanodine receptors

Gallstones can cause acute pancreatitis, an often fatal disease in which the pancreas digests itself. This is probably because of biliary reflux into the pancreatic duct and subsequent bile acid action on the acinar cells. Because Ca(2+) toxicity is important for the cellular damage in pancreatitis, we have studied the mechanisms by which the bile acid taurolithocholic acid 3-sulfate (TLC-S) liberates Ca(2+). Using two-photon plasma membrane permeabilization and measurement of [Ca(2+)] inside intracellular stores at the cell base (dominated by ER) and near the apex (dominated by secretory granules), we have characterized the Ca(2+) release pathways. Inhibition of inositol trisphosphate receptors (IP(3)Rs), by caffeine and 2-APB, reduced Ca(2+) release from both the ER and an acidic pool in the granular area. Inhibition of ryanodine receptors (RyRs) by ruthenium red (RR) also reduced TLC-S induced liberation from both stores. Combined inhibition of IP(3)Rs and RyRs abolished Ca(2+) release. RyR activation depends on receptors for nicotinic acid adenine dinucleotide phosphate (NAADP), because inactivation by a high NAADP concentration inhibited release from both stores, whereas a cyclic ADPR-ribose antagonist had no effect. Bile acid-elicited intracellular Ca(2+) liberation from both the ER and the apical acidic stores depends on both RyRs and IP(3)Rs.     

The distribution of the endoplasmic reticulum in living pancreatic acinar cells

Studies on pancreatic acinar cells provided the original evidence for the Ca(2+) releasing action of inositol 1,4,5-trisphosphate (IP(3)). Ironically, this system has presented problems for the general theory that IP(3) acts primarily on the endoplasmic reticulum (ER), because the IP(3)-elicited Ca(2+) release occurs in the apical pole, which is dominated by zymogen granules (ZGs) and apparently contains very little ER. Using confocal and two-photon microscopy and a number of different ER-specific fluorescent probes, we have now investigated in detail the distribution of the ER in living pancreatic acinar cells. It turns out that although the bulk of the ER, as expected, is clearly located in the baso-lateral part of the cell, there is significant invasion of ER into the granular pole and each ZG is in fact surrounded by strands of ER. This structural evidence from living cells, in conjunction with recent functional studies demonstrating the high Ca(2+) mobility in the ER lumen, provides the framework for a coherent and internally consistent theory for cytosolic Ca(2+) signal generation in the apical secretory pole, in which the primary Ca(2+) release occurs from ER extensions in the granular pole supplied with Ca(2+) from the main store at the base of the cell by the tunnel function of the ER.     

Local and global calcium signals and fluid and electrolyte secretion in mouse submandibular acinar cells

Polarized Ca(2+) signals that originate at and spread from the apical pole have been shown to occur in acinar cells from lacrimal, parotid, and pancreatic glands. However, "local" Ca(2+) signals, that are restricted to the apical pole of the cell, have been previously demonstrated only in pancreatic acinar cells in which the primary function of the Ca(2+) signal is to regulate exocytosis. We show that submandibular acinar cells, in which the primary function of the Ca(2+) signal is to drive fluid and electrolyte secretion, are capable of both Ca(2+) waves and local Ca(2+) signals. The generally accepted model for fluid and electrolyte secretion requires simultaneous Ca(2+)-activation of basally located K(+) channels and apically located Cl(-) channels. Whereas a propagated cell-wide Ca(2+) signal is clearly consistent with this model, a local Ca(2+) signal is not, because there is no increase in intracellular Ca(2+) concentration at the basal pole of the cell. Our data provide the first direct demonstration, in submandibular acinar cells, of the apical and basal location of the Cl(-) and K(+) channels, respectively, and confirm that local Ca(2+) signals do not Ca(2+)-activate K(+) channels. We reevaluate the model for fluid and electrolyte secretion and demonstrate that Ca(2+)-activation of the Cl(-) channels is sufficient to voltage-activate the K(+) channels and thus demonstrate that local Ca(2+) signals are sufficient to support fluid secretion.     

K+-induced ion-exchanges trigger trypsin activation in pancreas acinar zymogen granules

Trypsin premature activation has been thought to be a key event in the initiation phase of acute pancreatitis. Here we test a hypothesis that a sustained increase of cytosolic Ca(2+) concentration ([Ca(2+)](C)) can trigger K(+) influx into pancreas acinar zymogen granules (ZGs) via a Ca(2+)-activated K(+) channel (K(Ca)), and this influx of K(+) then mobilizes bound-Ca(2+) by K(+)/Ca(2+) ion-exchange to increase free Ca(2+) concentration in the ZGs ([Ca(2+)](G)) and release bound-H(+) by K(+)/H(+) ion-exchange to decrease the pH in ZGs (pH(G)). Both the increase of [Ca(2+)](G) and the decrease of pH(G) will facilitate trypsinogen autoactivation and stabilize active trypsin inside ZGs that could lead to acute pancreatitis. The experimental results are consistent with our hypothesis, suggesting that K(+) induced ion-exchanges play a critical role in the initiation of trypsin premature activation in ZGs.     

Perinuclear, perigranular and sub-plasmalemmal mitochondria have distinct functions in the regulation of cellular calcium transport

We have identified three distinct groups of mitochondria in normal living pancreatic acinar cells, located (i) in the peripheral basolateral region close to the plasma membrane, (ii) around the nucleus and (iii) in the periphery of the granular region separating the granules from the basolateral area. Three-dimensional reconstruction of confocal slices showed that the perigranular mitochondria form a barrier surrounding the whole of the granular region. Cytosolic Ca(2+) oscillations initiated in the granular area triggered mitochondrial Ca(2+) uptake mainly in the perigranular area. The most intensive uptake occurred in the mitochondria close to the apical plasma membrane. Store-operated Ca(2+) influx through the basolateral membrane caused preferential Ca(2+) uptake into sub-plasmalemmal mitochondria. The perinuclear mitochondria were activated specifically by local uncaging of Ca(2+) in the nucleus. These mitochondria could isolate nuclear and cytosolic Ca(2+) signalling. Photobleaching experiments indicated that different groups of mitochondria were not luminally connected. The three mitochondrial groups are activated independently by specific spatiotemporal patterns of cytosolic Ca(2+) signals and can therefore participate in the local regulation of Ca(2+) homeostasis and energy supply.     

Low extracellular pH induces damage in the pancreatic acinar cell by enhancing calcium signaling

Low extracellular pH (pHe) occurs in a number of clinical conditions and sensitizes to the development of pancreatitis. The mechanisms responsible for this sensitization are unknown. Because abnormal Ca(2+) signaling underlies many of the early steps in the pathogenesis of pancreatitis, we evaluated the effect of decreasing pHe from 7.4 to 7.0 on Ca(2+) signals in the acinar cell. Low pHe significantly increased the amplitude of cerulein-induced Ca(2+) signals. The enhancement in amplitude was localized to the basolateral region of the acinar cell and was reduced by pretreatment with ryanodine receptor (RYR) inhibitors. Because basolateral RYRs also have been implicated in the pathogenesis of pancreatitis, we evaluated the effects of RYR inhibitors on pancreatitis responses in acidic conditions. RYR inhibitors significantly reduced the sensitizing effects of low pHe on zymogen activation and cellular injury. These findings suggest that enhanced RYR-mediated Ca(2+) signaling in the basolateral region of the acinar cell is responsible for the injurious effects of low pHe on the exocrine pancreas.     

Phosphatidylinositol 3-kinase facilitates bile acid-induced Ca(2+) responses in pancreatic acinar cells

Bile acids are known to induce Ca(2+) signals in pancreatic acinar cells. We have recently shown that phosphatidylinositol 3-kinase (PI3K) regulates changes in free cytosolic Ca(2+) concentration ([Ca(2+)](i)) elicited by CCK by inhibiting sarco(endo)plasmic reticulum Ca(2+)-ATPase (SERCA). The present study sought to determine whether PI3K regulates bile acid-induced [Ca(2+)](i) responses. In pancreatic acinar cells, pharmacological inhibition of PI3K with LY-294002 or wortmannin inhibited [Ca(2+)](i) responses to taurolithocholic acid 3-sulfate (TLC-S) and taurochenodeoxycholate (TCDC). Furthermore, genetic deletion of the PI3K gamma-isoform also decreased [Ca(2+)](i) responses to bile acids. Depletion of CCK-sensitive intracellular Ca(2+) pools or application of caffeine inhibited bile acid-induced [Ca(2+)](i) signals, indicating that bile acids release Ca(2+) from agonist-sensitive endoplasmic reticulum (ER) stores via an inositol (1,4,5)-trisphosphate-dependent mechanism. PI3K inhibitors increased the amount of Ca(2+) in intracellular stores during the exposure of acinar cells to bile acids, suggesting that PI3K negatively regulates SERCA-dependent Ca(2+) reloading into the ER. Bile acids inhibited Ca(2+) reloading into ER in permeabilized acinar cells. This effect was augmented by phosphatidylinositol (3,4,5)-trisphosphate (PIP(3)), suggesting that both bile acids and PI3K act synergistically to inhibit SERCA. Furthermore, inhibition of PI3K by LY-294002 completely inhibited trypsinogen activation caused by the bile acid TLC-S. Our results indicate that PI3K and its product, PIP(3), facilitate bile acid-induced [Ca(2+)](i) responses in pancreatic acinar cells through inhibition of SERCA-dependent Ca(2+) reloading into the ER and that bile acid-induced trypsinogen activation is mediated by PI3K. The findings have important implications for the mechanism of acute pancreatitis since [Ca(2+)](i) increases and trypsinogen activation mediate key pathological processes in this disorder.     

Synthesis and degradation of fatty acid ethyl esters by cultured hepatoma cells exposed to ethanol

Fatty acid ethyl esters are a family of neutral lipids that are the products of esterification of fatty acids with ethanol. Unlike other pathways of ethanol metabolism, ethyl esters are present in numerous human organs which are the targets of ethanol-induced damage. In the present study, we have shown that fatty acid ethyl esters are synthesized by a hepatoma cell line in tissue culture when exposed to ethanol concentrations easily attained by man during social drinking. Unlike alcohol dehydrogenase, the enzyme(s) responsible for synthesis of ethyl esters are membrane-bound and concentrated in the microsomal fraction of rat hepatocytes. In addition, fatty acid ethyl esters are hydrolyzed to free fatty acids and ethanol by membrane-bound enzyme(s) that are enriched in the microsomal and mitochondrial-lysosomal fractions. Intracellular hydrolysis of fatty acid ethyl esters release free fatty acids which are preferentially incorporated into cellular cholesterol esters. Thus, we have shown that a hepatocellular line exposed to concentrations of ethanol easily achieved in man by social drinking utilize endogenous fatty acids to form long-lived ethanol metabolites, fatty acid ethyl esters. Importantly, this family of neutral lipids may act as biochemical mediators of ethanol-induced cell damage, including the changes in cholesterol metabolism noted in chronic alcoholics.     

Kinetic control of multiple forms of Ca(2+) spikes by inositol trisphosphate in pancreatic acinar cells.

The mechanisms of agonist-induced Ca(2+) spikes have been investigated using a caged inositol 1,4,5-trisphosphate (IP(3)) and a low-affinity Ca(2+) indicator, BTC, in pancreatic acinar cells. Rapid photolysis of caged IP(3) was able to reproduce acetylcholine (ACh)-induced three forms of Ca(2+) spikes: local Ca(2+) spikes and submicromolar (<1 microM) and micromolar (1-15 microM) global Ca(2+) spikes (Ca(2+) waves). These observations indicate that subcellular gradients of IP(3) sensitivity underlie all forms of ACh-induced Ca(2+) spikes, and that the amplitude and extent of Ca(2+) spikes are determined by the concentration of IP(3). IP(3)-induced local Ca(2+) spikes exhibited similar time courses to those generated by ACh, supporting a role for Ca(2+)-induced Ca(2+) release in local Ca(2+) spikes. In contrast, IP(3)- induced global Ca(2+) spikes were consistently faster than those evoked with ACh at all concentrations of IP(3) and ACh, suggesting that production of IP(3) via phospholipase C was slow and limited the spread of the Ca(2+) spikes. Indeed, gradual photolysis of caged IP(3) reproduced ACh-induced slow Ca(2+) spikes. Thus, local and global Ca(2+) spikes involve distinct mechanisms, and the kinetics of global Ca(2+) spikes depends on that of IP(3) production particularly in those cells such as acinar cells where heterogeneity in IP(3) sensitivity plays critical role.     

Bile acids induce a cationic current, depolarizing pancreatic acinar cells and increasing the intracellular Na+ concentration

Biliary disease is a major cause of acute pancreatitis. In this study we investigated the electrophysiological effects of bile acids on pancreatic acinar cells. In perforated patch clamp experiments we found that taurolithocholic acid 3-sulfate depolarized pancreatic acinar cells. At low bile acid concentrations this occurred without rise in the cytosolic calcium concentration. Measurements of the intracellular Na(+) concentration with the fluorescent probe Sodium Green revealed a substantial increase upon application of the bile acid. We found that bile acids induce Ca(2+)-dependent and Ca(2+)-independent components of the Na(+) concentration increase. The Ca(2+)-independent component was resolved in conditions when the cytosolic Ca(2+) level was buffered with a high concentration of the calcium chelator 1,2-bis(o-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA). The Ca(2+)-dependent component of intracellular Na(+) increase was clearly seen during stimulation with the calcium-releasing agonist acetylcholine. During acetylcholine-induced Ca(2+) oscillations the recovery of cytosolic Na(+) was much slower than the recovery of Ca(2+), creating a possibility for the summation of Na(+) transients. The bile-induced Ca(2+)-independent current was found to be carried primarily by Na(+) and K(+), with only small Ca(2+) and Cl(-) contributions. Measurable activation of such a cationic current could be produced by a very low concentration of taurolithocholic acid 3-sulfate (10 microm). This bile acid induced a cationic current even when applied in sodium- and bicarbonate-free solution. Other bile acids, taurochenodeoxycholic acid, taurocholic acid, and bile itself also induced cationic currents. Bile-induced depolarization of acinar cells should have a profound effect on acinar fluid secretion and, consequently, on transport of secreted zymogens.     

Calcium signalling and pancreatic cell death: apoptosis or necrosis?

Secretagogues, such as cholecystokinin and acetylcholine, utilise a variety of second messengers (inositol trisphosphate, cADPR and nicotinic acid adenine dinucleotide phosphate) to induce specific oscillatory patterns of calcium (Ca(2+)) signals in pancreatic acinar cells. These are tightly controlled in a spatiotemporal manner, and are coupled to mitochondrial metabolism necessary to fuel secretion. When Ca(2+) homeostasis is disrupted by known precipitants of acute pancreatitis, for example, hyperstimulation or non-oxidative ethanol metabolites, Ca(2+) stores (endoplasmic reticulum and acidic pool) become depleted and sustained cytosolic [Ca(2+)] elevations replace transient signals, leading to severe consequences. Sustained mitochondrial depolarisation, possibly via opening of the mitochondrial permeability transition pore (MPTP), elicits cellular ATP depletion that paralyses energy-dependent Ca(2+) pumps causing cytosolic Ca(2+) overload, while digestive enzymes are activated prematurely within the cell; Ca(2+)-dependent cellular necrosis ensues. However, when stress to the acinar cell is milder, for example, by application of the oxidant menadione, release of Ca(2+) from stores leads to oscillatory global waves, associated with partial mitochondrial depolarisation and transient MPTP opening; apoptotic cell death is promoted via the intrinsic pathway, when associated with generation of reactive oxygen species. Apoptosis, induced by menadione or bile acids, is potentiated by inhibition of an endogenous detoxifying enzyme NAD(P)H:quinone oxidoreductase 1 (NQO1), suggesting its importance as a defence mechanism that may influence cell fate.     

Stable Golgi-mitochondria complexes and formation of Golgi Ca(2+) gradients in pancreatic acinar cells

We have determined the localization of the Golgi with respect to other organelles in living pancreatic acinar cells and the importance of this localization to the establishment of Ca(2+) gradients over the Golgi. Using confocal microscopy and the Golgi-specific fluorescent probe 6-((N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)hexanoyl)sphingosine, we found Golgi structures localizing to the outer edge of the secretory granular region of individual acinar cells. We also assessed Golgi positioning in acinar cells located within intact pancreatic tissue using two-photon microscopy and found a similar localization. The mitochondria segregate the Golgi from lateral regions of the plasma membrane, the nucleus, and the basal part of the cytoplasm. The Golgi is therefore placed between the principal Ca(2+) release sites in the apical region of the cell and the important Ca(2+) sink formed by the peri-granular mitochondria. During acetylcholine-induced cytosolic Ca(2+) signals in the apical region, large Ca(2+) gradients form over the Golgi (decreasing from trans- to cis-Golgi). We further describe a novel, close interaction of the peri-granular mitochondria and the Golgi apparatus. The mitochondria and the Golgi structures form very close contacts, and these contacts remain stable over time. When the cell is forced to swell, the Golgi and mitochondria remain juxtaposed up to the point of cell lysis. The strategic position of the Golgi (closer to release sites than the bulk of the mitochondrial belt) makes this organelle receptive to local apical Ca(2+) transients. In addition the Golgi is ideally placed to be preferentially supplied by ATP from adjacent mitochondria.