Slow rather than fast calcium events encode physiological inputs and propagate within islets: Lessons from ultrafast imaging on acute pancreatic tissue slices

Pancreatic islets are micro-organs, mainly composed of insulin-secreting β cells, which play a central role in nutrient homeostasis and diabetes. They can be viewed as “mini-brains” of glucose homeostasis, as they present networks of excitable cells that express numerous neural proteins¹ and integrate nutritional, hormonal, and neuronal inputs in real time to continuously provide the amount of insulin required to cover physiological needs. In type 2 diabetes, which accounts for 90% of diabetes cases, both individual and collective β cell activities are impaired. Consequently, many groups have attempted to explore the single-cell and multicellular behavior of β cells for years using either intracellular electrophysiology, which offers high-temporal resolution but is invasive and limited to one cell, or optical methods, mainly Ca²⁺ imaging, which provides excellent spatial resolution but very limited temporal resolution, with a typical sampling rate of 0.5–2 Hz. This temporal resolution allows the detection of only slow Ca²⁺ events, namely Ca²⁺ bursts, and prevents the detection of fast Ca²⁺ events, namely Ca²⁺ spikes, although Ca²⁺ spikes represent the trigger for insulin granule exocytosis. In this issue of Acta Physiologica, Dolenšek et al. present high-temporal-resolution optical measurements (40–178 Hz) of selected islet areas using line scan confocal imaging on acute pancreas slices (i.e. in their native environment) in response to physiological levels of glucose and acetylcholine (Figure 1).² Their detailed characterization of both Ca²⁺ bursts and spikes at individual and collective levels offers new insights into the respective roles of these signals in islets, their encoding of glucose levels and cholinergic inputs, and their propagation within the micro-organ, and finally opens new perspectives for understanding islet “mini-brain” networks deregulation in diabetes.

The classical approach to measure spikes with sufficient temporal resolution in islets has so far been the perforated patch-clamp.³ However, this complex and invasive technique allows measurements at the single-cell level rather than the multicellular level and only for a few minutes, whereas islets are stimulated for 2–3h during digestion. Very few ultrafast Ca²⁺ measurements on isolated β cells⁴ or whole islets⁵ had been performed prior to this study, but they were limited in time and did not include spike analysis. Dolenšek et al. show now the correspondence between electrical and Ca²⁺ events,² consequently, their work paves the way for multicellular optical approaches as an alternative to patch-clamp for laboratories lacking the necessary equipment or expertise.

Their work provides new insights into islet biology at both cellular and multicellular levels. At the cellular level, the correlation between patch-clamp recordings and Ca²⁺ imaging here answers the controversial question of whether Ca²⁺ released from intracellular stores contributes to the glucose response.^{3, 6} Dolenšek et al.² show that this contribution is minimal and demonstrate that measuring electrical signals, including with extracellular electrodes,^7-9 accurately reflects intracellular Ca²⁺ dynamics. The new kinetic analysis of Ca²⁺ and electrical spikes here reveals that the falling phases and consequently the duration of Ca²⁺ spikes are slightly longer than those of electrical spikes (Figure 1).² This might be due to differences in kinetics between voltage-gated channels and the slower reduction of free Ca²⁺ through chelation by proteins and Ca²⁺ extrusion from the cytosol by ATPase pumps.

Ultrafast temporal resolution is not only necessary to study Ca²⁺ spikes but also for the precise analysis of Ca²⁺ bursts, as Dolenšek et al.² show that at least 2 Hz is required to avoid overestimation of burst durations. This minimum of 2 Hz has been used in many,^{10, 11} but unfortunately not all, previous studies measuring kinetic delays in Ca²⁺ responses between β cells to assess connectivity and propagation. This caution must be kept in mind for the design and interpretation of imaging experiments.

The data of Dolenšek et al.² on the progressive decrease in spike frequency during Ca²⁺ bursts (Figure 1) confirm previous findings obtained with both intracellular⁵ and extracellular electrophysiology⁸: action potentials are more frequent in the first half of the slow electrical events than in their second half. As discussed by the authors, this is likely due in large part to Ca²⁺-activated K⁺ (K_Ca) channels and to the reopening of some K_ATP channels via a negative Ca²⁺ feedback on the metabolism and ATP consumption by ATPase pumps.³ This composite K⁺ current has been named K_slow and may act as a pacemaker in β cells, although additional mechanisms cannot be excluded.

A major conclusion of this study is that glucose levels are encoded by Ca²⁺ burst frequency and active time and not by spikes,² in accordance with previous electrophysiological data.⁸ The multicellular analysis on line scans reveals that spikes propagate faster than bursts, have little intercellular coactivity, and remain relatively localized, whereas bursts exhibit the highest coactivity and spread across the islet (Figure 1).² In addition, coactivity and propagation velocity of bursts, but not of spikes, depend on glucose levels. A glucose-dependent increase in gap junctions upon prolonged glucose stimulation¹² could contribute to the glucose effect on Ca²⁺ propagation. Collectively, bursts are therefore slow multicellular events that both drive and carry the spikes, which are unicellular events. Thus, the electrical slow potentials (SPs) propagating within islets and recorded previously using high-density multielectrode arrays⁷ clearly correspond to the multicellular Ca²⁺ bursts. Similarly to the SPs,⁷ and as previously reported for Ca²⁺ dynamics in pancreas slices,¹³ the intercellular coactivity of bursts is low at the beginning of the glucose response, corresponding to weak β cell coupling in the first phase, and then considerably increases in the second phase.² Interestingly, the fact that acetylcholine modulates bursts but not spikes² (Figure 1) suggests that this neurotransmitter primarily influences collective rather than individual β cell activity.

In the future, it would be interesting to dissect the role of slow and fast Ca²⁺ signals in insulin secretion kinetics, potentially by combining the method of Dolenšek et al.² and measurements of vesicle fusion using TIRF microscopy.¹⁴ Examining the effects of other physiologically relevant inputs, such as amino acids and gut-derived incretins, on slow and fast signal propagations could complete the model. The combination of this approach with microfluidics may enable the application of gradual nutrient increase and decrease kinetics observed in vivo.¹⁵ Importantly, human islets differ from mouse islets in terms of proportion and spatial distribution of β and non-β cells, as well as at the cellular level since human β cells express more fast-kinetics Na⁺ and T-type Ca²⁺ channels and more K_Ca channels.³ This would likely impact the organization of burst and spike signals and their propagation in human islets. In addition, it will be important to determine whether non-β cells contribute to or modulate the coactivities reported in this study. Finally, future technical improvements may allow whole islet capture and prevent the inherent photobleaching that limits long-term (>30 min) evaluations. In this perspective, the approach could be complemented by high-resolution and minimally invasive electrophysiological approaches in 2D⁷ or even in 3D⁹ to better understand the fascinating “mini-brain” islet microorgans and to elucidate how electrical and Ca²⁺ signal dynamics and propagation are altered in metabolic disorders such as type 2 diabetes.

Matthieu Raoux: Writing – original draft; writing – review and editing; conceptualization; supervision. Dorian Chapeau: Writing – review and editing. Jochen Lang: Writing – review and editing.