Movement of calcium signals and calcium-binding proteins: firewalls

Non-Vesicular Intracellular Traffic

Movement of calcium signals and calcium-binding proteins: firewalls, traps and tunnels S.L. Barrow, M.W. Sherwood, N.J. Dolman1 , O.V. Gerasimenko, S.G. Voronina and A.V. Tepikin2 MRC Secretory Control Research Group, The Physiological Laboratory, The University of Liverpool, Liverpool L69 3BX, U.K.

Abstract In the board game ‘Snakes and Ladders’, placed on the image of a pancreatic acinar cell, calcium ions have to move from release sites in the secretory region to the nucleus. There is another important contraflow – from calcium entry channels in the basal part of the cell to ER (endoplasmic reticulum) terminals in the secretory granule region. Both transport routes are perilous as the messenger can disappear in any place on the game board. It can be grabbed by calcium ATPases of the ER (masquerading as a snake but functioning like a ladder) and tunnelled through its low buffering environment, it can be lured into the whirlpools of mitochondria uniporters and forced to regulate the tricarboxylic acid cycle, and it can be permanently placed inside the matrix of secretory granules and released only outside the cell. The organelles could trade calcium (e.g. from the ER to mitochondria and vice versa) almost depriving this ion the light of the cytosol and noble company of cytosolic calcium buffers. Altogether it is a rich and colourful story.

The activation of calcium signals in pancreatic acinar cells Local calcium transients (Figure 1A) are the main form of calcium signalling in pancreatic acinar cells. These transients preferentially occur in the part of the cell containing secretory granules [1,2]. Local calcium signals are perfectly capable of activating secretion of digestive enzymes (by exocytosis) and activating fluid secretion [2]. This is clearly an economical and efficient way of signalling, occurring just where it is needed. There are two main hypotheses that can explain the formation of these transients: (i) the transients occur because the receptors of calcium releasing agonists are close to the secretory granule region and this region simply develops the highest concentration of second messengers; (ii) receptors are located on the basal membrane and the second messengers (e.g. inositol 1,4,5-trisphosphate) are produced in the basal region of the cell, but the most sensitive calcium releasing mechanisms are localized to the secretory granule region. The second hypothesis implies the long distance translocation of second messengers. To distinguish between these two possibilities, we developed and applied a novel technique: intra-patch pipette uncaging [3]. The idea behind this technique was to occlude and stimulate only a specific very small region of the cell membrane. The cell-attached patch clamp configuration allows this very specific and local stimulation. To activate receptors in the patch, limited by the boundaries of glass–lipid contact, we used uncaging of caged CCh Key words: acetylcholine, calcium signalling, calcium-binding protein, calmodulin, mitochondrion, pancreatic acinar cell. Abbreviations used: ACh, acetylcholine; CCh, carbachol; CCK, cholecystokinin; ER, endoplasmic reticulum. 1 Present address: Synaptic Physiology Unit, NINDS (National Institute of Neurological Disorders and Stroke), Porter Neuroscience Research Center, Building 35, Room 3C-1004, 35 Convent Drive, MSC 3701, National Institutes of Health, Bethesda, MD 20892-3701, U.S.A. To whom correspondence should be addressed (email [email protected]).

2

(carbachol). We found [3] that when the small region in the basal membrane was stimulated (by directing UV laser light to the pipette and uncaging the CCh), the calcium response developed on the opposite site of the cell – in the granule region (more than 15 µm from the site of stimulation). These experiments strongly favour long distance translocation of the second messenger(s) and preferential high sensitivity of calcium release mechanisms, located in the secretory granule region of the cell. A by-product of these experiments was the realization of how powerful the amplification mechanism is in this signalling cascade – stimulation of only 1–2% of the cell surface was sufficient to trigger local apical calcium signals, global calcium signals that originated in the apical and propagate to the basal region of the cell and even calcium responses propagating from one cell to another. Importantly, we were recently able to resolve both local apical calcium signals and global calcium transients (initiated in the apical region) in intact pancreatic acinar cells, located within undissociated pancreatic tissue [3].

The organelles in the firing line The basal part of cytosol (and particularly the nucleus) is very well protected from a calcium increase during local apical calcium transients [4]. Very impressive calcium gradients up to 500 nM/µm form between apical and basal parts of pancreatic acinar cells. These standing gradients can be remarkably long-lasting: secretagogues can produce local calcium transients that last more than 10 s [4]; bile acids (considered as potential triggers of acute pancreatitis) can activate local calcium signals outlasting 20 s [5]. Calcium released in the apical part of an acinar cell faces a hostile environment. This cell type has a very high calcium-binding capacity of the cytosol – approx. 1500 [6], which befits a small cell aiming to display local calcium signals. Calcium can be also effectively extruded by calcium pumps which are preferentially
C 2006

Biochemical Society

381

382

Biochemical Society Transactions (2006) Volume 34, part 3

Figure 1 ACh-induced calcium signals and localization of organelles in pancreatic acinar cells (A) Left image shows local calcium signals in the doublet of pancreatic acinar cells at the peak of its development (recorded using Fura Red; here and in the right image, red and yellow correspond to high calcium concentration, and green and blue correspond to low calcium concentration); right image depicts a calcium wave originated in the secretory granule regions at the moment of its spreading to the basal regions of the cells. Scale bar, 5 µm. Reproduced from Pflugers ¨ Archiv European Journal of Physiology, 432, 1996, pp. 1055–1061, ‘Short pulses of acetylcholine stimulation induce cytosolic Ca2+ signals that are excluded from the nuclear region in pancreatic acinar cells’ by O.V. Gerasimenko, J.V. Gerasimenko, O.H. Petersen and A.V. and Tepikin, Figure 2, with c 1996 kind permission of Springer Science and Business Media.
Springer-Verlag. (B) Transmitted light image of a cluster of pancreatic acinar cells (left image). The right image shows the distribution of Golgi [stained with NBD (7-nitrobenz-2-oxa-1,3-diazole) C6 -ceramide, shown in green] and mitochondria (shown in red, revealed with MitoTracker Deep Red). Scale bar, 5 µm. Reprinted from [10] with permission. c 2005 American Society for Biochemistry and Molecular Biology.


localized to the apical membrane of the cells [7–9]. But calcium buffer alone cannot explain the formation of very long-lasting calcium transients on the boundary between apical and basal regions of the cell. And calcium pumps of the plasma membrane cannot prevent calcium transients from spreading to the basal part of the cell. There should be calcium-transporting (and storing) organelles responsible for such remarkable apical/basal calcium gradients. What are the organelles affected by the local apical calcium signals? The first target is secretory granules. The whole secretory granule region is exposed to a calcium increase if the Ca2+ transient is longer than 2–3 s. Important secretagogues such as the circulatory hormone CCK (cholecystokinin) and neurotransmitter ACh (acetylcholine) employ Ca2+ signals to
C 2006

Biochemical Society

induce and regulate secretion. After secretory granules, the next prominent organelle found on a line from the apex of the cell to the basal membrane is the Golgi apparatus (Figure 1B). We used fluorescent tracers to visualize calcium signals and the position of cellular organelles simultaneously [10]. This technology allowed us to put calcium signals ‘on the map’ of cellular organelles. It was clear from our experiments that during local calcium transients the Golgi is exposed to calcium signals and a calcium gradient is formed between cis and trans components of this organelle [10]. The major group of mitochondria (termed ‘mitochondrial belt’) is positioned next to the Golgi (Figure 1B). The two organelles form very close contacts (the boundary of organelles are indistinguishable on high-resolution confocal images). The mitochondrial belt also projects along the lateral membrane of the pancreatic acinar cell. Therefore the Golgi is surrounded by mitochondria from both lateral and basal sites. Profiling of calcium signals and distribution of fluorescence of the mitochondrial probe indicate that calcium signals dissipate within the mitochondrial belt [10]. This reinforces the notion that mitochondria play an important role in preventing calcium signals from entering the basal part of the cell – a phenomenon discovered previously in our laboratory [11,12]. The ‘mitochondrial firewall’ (a term introduced by Dr David Yule from University of Rochester, U.S.A.) not only shapes calcium signals but the position of mitochondria is also important for the signalling–metabolism coupling in pancreatic acinar cells. This was nicely illustrated by simultaneous monitoring of calcium signals and the concentration of mitochondrial NADH [13]. The concentration of this important reducing equivalent undergoes transient elevations following calcium signals. The minimum duration of calcium signals necessary to induce measurable NADH response was approx. 2.5 s [13]. For the short cytosolic calcium transients, the duration of NADH response was much longer than the duration of the calcium signal that triggered the response. This suggests a very clever ‘proactive’ mechanism of energy production – the extended acceleration of the rate of tricarboxylic acid cycle by calcium-dependant dehydrogenases should allow for restoration of ionic gradients that occur as a consequence of calcium responses. It is particularly important since this cell type contains powerful calcium-dependant channels that change the cytosolic concentration of chloride, sodium and potassium during stimulation. For example, we have recently shown that during a single ACh-induced calcium transient, cytosolic sodium concentration changes by more than 5 mM [14]. The effects of calcium signals on mitochondria are also evident in experiments with measurements of the mitochondria membrane potential with TMRM (tetramethylrhodamine methyl ester) or TMRE (tetramethylrhodamine ethyl ester). We found that physiological doses of calcium-releasing secretagogues can trigger the depolarization of mitochondria. Interestingly, this secretagoguedependant depolarization was resolved only using quench– dequench mode of measurements (in these experiments, mitochondria are loaded with very high concentrations of

Non-Vesicular Intracellular Traffic

the indicator). This is not surprising since our preliminary experiments (conducted by S.L. Barrow) showed that the quench–dequench mode of measurement is approximately two orders of magnitude more sensitive than the low concentration mode [15]. Although the local apical calcium transients are probably the main form of calcium responses in pancreatic acinar cells, the physiological concentrations of CCK will also produce global calcium signals. During such global signals, the calcium increase starts in the apical part of the cell (Figure 1A) and then propagates to the basal region. The powerful calcium-induced calcium-release mechanism helps to overwhelm the mitochondrial firewall and bring Ca2+ signals into the basal region of the cell (containing a high density of rough ER) and into the nucleus [16]. Local calcium signals and global calcium signals have very different effects on the distribution of an abundant and important calcium target – calmodulin. Short calcium responses trigger calmodulin translocation into the secretory granule region of the acinar cells, whereas more substantial transients induce nuclear transport of calmodulin and oscillations of nuclear calmodulin concentration [17]. The nuclear accumulation of calmodulin occurs slowly, but the loss of calmodulin from the nucleus is an even slower process, therefore the nucleus works as an integrator (trap) of calmodulin – summating pulses of calmodulin translocation occurring due to individual global calcium transients [17].

Calcium store and its replenishment In the apical part of the acinar cell, the volume of strands of the ER is small [10,18]. The ER serves as the main (but probably not the only) calcium store in pancreatic acinar cells. And yet its density is very low in the part of the cell that displays local calcium signals and serves as the initiation region for global calcium waves. Even more interesting is the fact that calcium oscillations in the apical part of the cell can last for a considerable period of time in cells placed in calcium-free extracellular solution. In other words, tiny projections of ER in the apical part of pancreatic acinar cells behave as if they are almost undepletable. What can explain this remarkable behaviour? Firstly, insights into the possible mechanism came from a study conducted using a local Ca2+ reloading technique [19]. In the beginning of the experiment, the calcium stores were depleted by applying supramaximal concentration of the calcium-releasing agonist ACh to cells based in calciumfree extracellular solution. Then the agonist was removed and refilling was allowed to occur, but only via a small patch of the basal membrane, limited by the patch pipette in cell-attached configuration. The pipette contained a high calcium concentration (usually 10 mM), while the rest of the extracellular solution was calcium-free. When the cell was restimulated after the period of loading via the basal patch, the calcium responses were initiated in the apical part of the cell. Importantly, the local reloading occurred without raising the cytosolic calcium level. The only organelle linking basal and apical parts of the cell is ER. Overall the experiments

led to formulation of the tunnelling hypothesis – calcium entering via the basal membrane is taken into the lumen of the ER and then transported via the lumen of this organelle to its release sites in the apical region of the cell [19]. The term ‘tunnelling’, as a description of this phenomenon, was coined by Ole H. Petersen. Later the movement of fluorescent probes and calcium via ER lumen was directly visualized by Myoung Park in our laboratory [18]. The tiny projections of ER in the apical part of the cells are connected to the bulk of extremely well-developed ER in the basal region and can import calcium by a fast tunnelling process. This is indeed a remarkably clever space-saving solution for a busy exocrine secretory cell. An interesting question – How is the resting calcium level maintained in the ER lumen? – was addressed in our laboratory using direct calcium measurements in the ER lumen using low-affinity calcium indicators. Utilizing this approach, we characterized the calcium dependency of calcium uptake and calcium leak [20]. We also found an important role of protein-conducting channel translocons in mediating calcium leak in non-stimulated cells [21].

Concluding remarks Calcium signalling is a mature field with a well-developed methodology and considerable depth of understanding of mechanisms and downstream reactions. An important line of development of the calcium signalling field is the application of this fundamental knowledge to the elucidation of the mechanisms of calcium-dependant pathology. In the case of exocrine pancreas, a specific question is the involvement of calcium in an important disease – acute pancreatitis. Some advances in this direction have already been made [5,8,22,23] but this work is certainly at an early developmental stage. Work in our laboratory is supported by an MRC programme grant. S.L.B. is an MRC-funded Ph.D. student. M.W.S. is a Wellcome Trustfunded Ph.D. student. We are grateful to Mr M. Houghton for excellent technical assistance.

References 1 Ashby, M.C. and Tepikin, A.V. (2002) Physiol. Rev. 82, 701–734 2 Petersen, O.H. (1992) J. Physiol. 448, 1–51 3 Ashby, M.C., Camello-Almaraz, C., Gerasimenko, O.V., Petersen, O.H. and Tepikin, A.V. (2003) J. Biol. Chem. 278, 20860–20864 4 Gerasimenko, O.V., Gerasimenko, J.V., Petersen, O.H. and Tepikin, A.V. (1996) Pflugers Arch. 432, 1055–1061 5 Voronina, S., Longbottom, R., Sutton, R., Petersen, O.H. and Tepikin, A. (2002) J. Physiol. 540, 49–55 6 Mogami, H., Gardner, J., Gerasimenko, O.V., Camello, P., Petersen, O.H. and Tepikin, A.V. (1999) J. Physiol. 518, 463–467 7 Belan, P., Gerasimenko, O., Petersen, O.H. and Tepikin, A.V. (1997) Cell Calcium 22, 5–10 8 Lee, M.G., Xu, X., Zeng, W., Diaz, J., Kuo, T.H., Wuytack, F., Racymaekers, L. and Muallem, S. (1997) J. Biol. Chem. 272, 15771–15776 9 Tepikin, A.V., Llopis, J., Snitsarev, V.A., Gallacher, D.V. and Petersen, O.H. (1994) Pflugers Arch. 428, 664–670 10 Dolman, N.J., Gerasimenko, J.V., Gerasimenko, O.V., Voronina, S.G., Petersen, O.H. and Tepikin, A.V. (2005) J. Biol. Chem. 280, 15794–15799 11 Park, M.K., Ashby, M.C., Erdemli, G., Petersen, O.H. and Tepikin, A.V. (2001) EMBO J. 20, 1863–1874
C 2006

Biochemical Society

383

384

Biochemical Society Transactions (2006) Volume 34, part 3

12 Tinel, H., Cancela, J.M., Mogami, H., Gerasimenko, J.V., Gerasimenko, O.V., Tepikin, A.V. and Petersen, O.H. (1999) EMBO J. 18, 4999–5008 13 Voronina, S., Sukhomlin, T., Johnson, P.R., Erdemli, G., Petersen, O.H. and Tepikin, A. (2002) J. Physiol. 539, 41–52 14 Voronina, S.G., Gryshchenko, O.V., Gerasimenko, O.V., Green, A.K., Petersen, O.H. and Tepikin, A.V. (2005) J. Biol. Chem. 280, 1764–1770 15 Voronina, S.G., Barrow, S.L., Gerasimenko, O.V., Petersen, O.H. and Tepikin, A.V. (2004) J. Biol. Chem. 279, 27327–27338 16 Ashby, M.C., Craske, M., Park, M.K., Gerasimenko, O.V., Burgoyne, R.D., Petersen, O.H. and Tepikin, A.V. (2002) J. Cell Biol. 158, 283–292 17 Craske, M., Takeo, T., Gerasimenko, O., Vaillant, C., Torok, K., Petersen, O.H. and Tepikin, A.V. (1999) Proc. Natl. Acad. Sci. U.S.A. 96, 4426–4431


C 2006

Biochemical Society

18 Park, M.K., Petersen, O.H. and Tepikin, A.V. (2000) EMBO J. 19, 5729–5739 19 Mogami, H., Nakano, K., Tepikin, A.V. and Petersen, O.H. (1997) Cell 88, 49–55 20 Mogami, H., Tepikin, A.V. and Petersen, O.H. (1998) EMBO J. 17, 435–442 21 Lomax, R.B., Camello, C., Van Coppenolle, F., Petersen, O.H. and Tepikin, A.V. (2002) J. Biol. Chem. 277, 26479–26485 22 Pandol, S.J. (2005) Curr. Opin. Gastroenterol. 21, 538–543 23 Raraty, M., Ward, J., Erdemli, G., Vaillant, C., Neoptolemos, J.P., Sutton, R. and Petersen, O.H. (2000) Proc. Natl. Acad. Sci. U.S.A. 97, 13126–13131 Received 14 December 2005