SummaryWe have studied how the F-actin cytoskeleton is involved in establishing the heterogeneous intracellular Ca 2+ levels ([Ca 2+ ] i ) and in the organization of the exocytotic machinery in cultured bovine chromaffin cells. Simultaneous confocal visualization of [Ca 2+ ] i and transmitted light studies of the cytoskeleton showed that, following cell stimulation, the maximal signal from the Ca 2+ -sensitive fluorescent dye Fluo-3 was in the empty cytosolic spaces left by cytoskeletal cages. This was mostly due to the accumulation of the dye in spaces devoid of cytoskeletal components, as shown by the use of alternative Ca 2+ -insensitive fluorescent cytosolic markers. In addition to affecting the distribution of such compounds in the cytosol, the cytoskeleton influenced the location of L-and P-Q-type Ca 2+ channel clusters, which were associated with the borders of cytoskeletal cages in resting and stimulated cells. Indeed, syntaxin-1 and synaptotagmin-1, which are components of the secretory machinery, were present in the same location. Furthermore, granule exocytosis took place at these sites, indicating that the organization of the F-actin cytoskeletal cortex shapes the preferential sites for secretion by associating the secretory machinery with preferential sites for Ca 2+ entry. The influence of this cortical organization on the propagation of [Ca 2+ ] i can be modelled, illustrating how it serves to define rapid exocytosis.

Mitochondria play a significant role in shaping cytosolic Ca signals. Thus, transfer of Ca across the mitochondrial membrane is much studied, not only in intact cells but also in artificial systems such as mitochondrial suspensions or permeabilised cells. Observed rates of Ca changes vary by at least one order of magnitude. In this work, we investigate the relationship between the Ca dynamics observed in various experimental conditions using a computational model calibrated on experimental data. Results confirm that mitochondrial Ca exchange fluxes through the mitochondrial Ca uniporter (MCU) and the Na /Ca exchanger obey the same basic kinetics in cells and in suspensions, and emphasise the important role played by the high Ca levels reached in mitochondria-associated endoplasmic reticulum membranes in intact cells. Tissue specificity can be ascribed to the different modes of regulation of the MCU by Ca , probably related to the specific levels of expression of the Ca sensing regulator subunit of this channel. The model emphasises the importance of mitochondrial density and buffering in controlling the rate of Ca exchanges with mitochondria, as verified experimentally. Finally, we show that heterogeneity between individual mitochondria can explain the large range of amplitudes and rates of rise in mitochondrial Ca concentration that have been observed experimentally.

(2010) The organization of the secretory machinery in chromaffin cells as a major factor in modeling exocytosis, HFSP Journal, 4:2, 85-92

Glucagon hormone is synthesized and released by pancreatic α-cells, one of the islet-cell types. This hormone, along with insulin, maintains blood glucose levels within the physiological range. Glucose stimulates glucagon release at low concentrations (hypoglycemia). However, the mechanisms involved in this secretion are still not completely clear. Here, using experimental calcium time series obtained in mouse pancreatic islets at low and high glucose conditions, we propose a glucagon secretion model for α-cells. Our model takes into account that the resupply of releasable granules is not only controlled by cytoplasmic , as in other neuroendocrine and endocrine cells, but also by the level of extracellular glucose. We found that, although calcium oscillations are highly variable, the average secretion rates predicted by the model fall into the range of values reported in the literature, for both stimulated and non-stimulated conditions. For low glucose levels, the model predicts that there would be a well-controlled number of releasable granules refilled slowly from a large reserve pool, probably to ensure a secretion rate that could last for several minutes. Studying the α-cell response to the addition of insulin at low glucose, we observe that the presence of insulin reduces glucagon release by decreasing the islet level. This observation is in line with previous work reporting that dynamics, mainly frequency, is altered by insulin [1]. Thus, the present results emphasize the main role played by and glucose in the control of glucagon secretion by α-cells. Our modeling approach also shows that calcium oscillations potentiate glucagon secretion as compared to constant levels of this cellular messenger. Altogether, the model sheds new light on the subcellular mechanisms involved in α-cell exocytosis, and provides a quantitative predictive tool for studying glucagon secretion modulators in physiological and pathological conditions.

In chromaffin cells, SNARE proteins, forming the basic exocytotic machinery are present in membrane clusters of 500-600 nm in diameter. These microdomains containing both SNAP-25 and syntaxin-1 are dynamic and the expression of altered forms of SNAREs modifies not only their motion but also the mobility of the associated granules. It is also clear that SNARE microdomain location defines the place for individual vesicle fusion and that the alteration of cluster dynamics affects the fusion process itself. Interestingly, these SNARE patches colocalize with the borders of F-actin cages forming the cytoskeletal cortical network, and these borders also contain clusters of L- and P/Q type calcium channels. The organization of the secretory machinery in association with the borders of cytoskeletal cages seems to be an effective way to promote fast coupling between calcium entry and catecholamine release as demonstrated with the use of mathematical secretory models.

No abstract

No abstract

Minimal mathematical models were developed to describe the electrophysiological properties of human δ‐cells. Markov models of single channels were first developed based on the analysis of electrophysiological data. Monte Carlo simulations of voltage‐clamp experiments were performed in an iteratively optimization procedure to estimate the number of channels required to reproduce the main characteristics of the macroscopic currents recorded experimentally. A membrane model of the firing of action potentials was then developed based on the kinetic schemes of single channels and the number of channels estimated. We showed that macroscopic currents of human δ‐cells can be reproduced by minimal models of single channels when the appropriate number of channels is considered. In addition, our simulations suggest that human δ‐cells are capable of generating action potentials through the interaction of the ionic currents involved. Finally, we determined the relative contribution of the currents underlying the firing of action potentials in human pancreatic δ‐cells, which allowed us to propose a qualitative model of an action potential in terms of the underlying ionic currents.