Calcium is the undisputed trigger of insulin release. Extracellular ionized calcium sits at roughly 1.1 to 1.3 millimolar, about ten thousand times higher than the cytosolic concentration near 100 nanomolar, and beta cells exploit this steep gradient with exquisite precision. When blood glucose rises, glucose enters the cell through transporters—GLUT2 in rodents, the higher-affinity GLUT1 in humans—and is funneled almost entirely into mitochondrial oxidation rather than lactate production. The resulting surge in the ATP-to-ADP ratio closes ATP-sensitive potassium channels, depolarizes the plasma membrane, and opens voltage-gated calcium channels. The incoming calcium binds synaptotagmin on secretory granules, engaging the SNARE fusion machinery and driving insulin exocytosis in the characteristic biphasic pattern seen in both humans and rodents.

The channels themselves form a layered hierarchy. L-type channels dominate: in mice, beta-cell-specific deletion of CaV1.2 nearly abolishes first-phase insulin secretion and causes glucose intolerance, while CaV1.3 contributes to electrical activity initiation and, notably, may be expressed tenfold higher than CaV1.2 in human beta cells, where its genetic variants associate with type 2 diabetes. Transient receptor potential channels such as TRPM2, which is activated by hydrogen peroxide and intracellular calcium, and TRPM5 and TRPV2, act as amplifiers and fine-tuners rather than primary triggers. Equally critical are the extrusion systems—the plasma membrane calcium ATPase and the sodium-calcium exchanger—which clear cytosolic calcium and, when heterozygously deleted in mice, paradoxically enhance glucose-stimulated insulin secretion by expanding intracellular stores.

Inside the cell, two organelles dominate calcium storage. The endoplasmic reticulum combines high capacity with high affinity, holding calcium at roughly 200 to 500 micromolar through the SERCA pumps, with SERCA2b and SERCA3 predominant in beta cells. Release occurs through inositol trisphosphate receptors, activated by PLC-generated IP3 from G protein-coupled receptor signaling, and through ryanodine receptors, which mediate calcium-induced calcium release. Mitochondria, by contrast, take up calcium with low affinity but enormous capacity, relying on the mitochondrial calcium uniporter and the high local calcium concentrations generated at mitochondria-associated membranes, the contact sites where IP3 receptors, the chaperone Grp75, and the channel VDAC1 assemble into a functional transfer complex. Matrix calcium activates TCA cycle dehydrogenases, boosting NADH and FADH2 production and ATP synthesis—the metabolic signal that completes stimulus-secretion coupling. Efflux through the NCLX exchanger is equally essential: its silencing in mouse models causes mitochondrial calcium accumulation and delayed, impaired insulin secretion.

Reactive oxygen species speak a different but equally structured language. Superoxide anions, generated as byproducts of the mitochondrial electron transport chain or deliberately by NADPH oxidase enzymes, are dismutated by SOD1 and SOD2 into hydrogen peroxide, the most signaling-competent member of the family. Hydrogen peroxide is stable enough to diffuse through peroxiporin aquaporins such as AQP8 yet reactive enough to oxidize cysteine residues on target proteins reversibly, forming sulfenic acid. In beta cells, this chemistry directly touches the secretory machinery: oxidation of the KATP channel’s SUR1 regulatory subunit enhances channel closure, reinforcing depolarization and calcium influx. Pharmacological or genetic suppression of NOX enzymes reduces glucose-stimulated insulin secretion in rat islets and human islets alike, and NOX4, which constitutively produces hydrogen peroxide coupled to glucose metabolism, has emerged as a required signal—its deletion abolishes first-phase secretion in mouse islets and INS-1E cells.

Recent subcellular mapping in rat INS-1E beta cells reveals that NOX isoforms are not randomly distributed. NOX1 and DUOX1 concentrate in the endoplasmic reticulum, DUOX2 localizes to insulin-containing vesicles, and NOX2 and NOX4 span vesicles, ER, and the plasma membrane. This compartmentalization creates spatially restricted redox microdomains in which hydrogen peroxide selectively modifies nearby targets—ER luminal redox balance for protein folding, granule-associated proteins for insulin processing, and ion channels at the membrane for excitability—rather than flooding the cell with indiscriminate oxidation. Pro-inflammatory cytokines remodel this landscape in a temporally coordinated fashion, suggesting a shift from controlled signaling toward sustained oxidative conditions during beta-cell stress.

The deepest integration occurs at mitochondria-associated membranes. There, calcium released from the ER through IP3 receptors generates localized hotspots that drive mitochondrial calcium uptake via the uniporter, stimulating oxidative phosphorylation and, consequently, mitochondrial ROS production. The crosstalk is bidirectional: IP3 receptors and VDAC are themselves redox-sensitive, with moderate oxidation enhancing channel activity while excessive oxidation impairs function and disrupts calcium homeostasis. Tethering proteins such as mitofusin 2 maintain the physical proximity that makes these microdomains possible. In beta cells, whose antioxidant profile includes robust peroxiredoxin and thioredoxin systems but relatively low catalase and glutathione peroxidase, this architecture allows efficient handling of transient physiological ROS while leaving limited reserve for chronic stress.

Type 2 diabetes progressively dismantles this coordination. In obesity and insulin resistance, beta cells initially compensate with enhanced calcium signaling and insulin output, but the compensation is fragile. High-fat diet feeding eventually uncouples voltage-gated calcium entry from the exocytotic machinery, and loss of calcium coordination between individual beta cells precedes overt dysfunction. Palmitate exposure depletes ER calcium through reduced SERCA2b expression, elevates cytosolic and mitochondrial calcium, and generates ER stress. Human islets from donors with type 2 diabetes show reduced ER-mitochondria contacts and impaired calcium transfer, while chronic glucotoxicity disrupts ER-mitochondrial calcium exchange even as organelle interactions increase. Mitochondrial ROS overproduction, driven by electron transport chain dysfunction and amplified by calcium overload, feeds forward into further calcium dysregulation through redox-sensitive TRPM2 channels and impaired IP3 receptor and ryanodine receptor function.

The redox side of the disease is equally layered. Sustained glucolipotoxicity and inflammation elevate ROS from mitochondria, NOX1 and NOX2, peroxisomes, and the ER simultaneously. NOX1 expression rises in islets from individuals with type 2 diabetes, and its cytokine-driven activation abolishes glucose-stimulated insulin secretion—an effect prevented by NOX1 inhibition. NOX2-derived hydrogen peroxide mediates palmitate-induced lipotoxicity, and ER stress can precede NOX2 activation, suggesting a temporal hierarchy in which ER dysfunction triggers ROS amplification. Antioxidant defenses, dependent on NADPH availability through the pentose phosphate pathway and anaplerotic metabolism, become overwhelmed: chronic high glucose reduces glucose-6-phosphate dehydrogenase, weakening NADPH supply, while overexpression of G6PD or ER-resident glutathione peroxidases restores function under stress. Emerging evidence even implicates sustained ROS in basal insulin hypersecretion before overt secretory failure, depleting insulin stores and accelerating beta-cell exhaustion.

Beyond individual cells, the islet operates as a coordinated network. Gap junctions composed of connexin 36 synchronize calcium oscillations across beta cells, and their disruption abolishes pulsatile insulin secretion. Specialized hub cells and wave-initiator cells organize propagation, and paracrine signals from glucagon, somatostatin, and urocortin 3 fine-tune the network. Redox signaling participates in this coordination too: hydrogen peroxide can diffuse between cells, and connexin 36 itself is sensitive to cysteine oxidation, meaning local redox tone may shape calcium wave propagation. In type 2 diabetes, calcium wave propagation, functional connectivity, and global synchronization all decline. The review’s authors argue that future progress demands genetically encoded biosensors for calcium and hydrogen peroxide targeted to specific organelles, applied to human islets, to resolve the temporal sequence linking ROS generation and calcium dysregulation. Understanding this multi-organellar dialogue, they contend, may unlock pharmacological strategies to protect beta-cell function or promote functional recovery before irreversible failure sets in.

Subject of Research: Organelle-specific crosstalk between calcium and redox signaling in pancreatic beta cells in physiology and type 2 diabetes

Article Title: Organelle-specific crosstalk between calcium and redox signaling in pancreatic β-cells in physiology and type 2 diabetes

Article References: Camara, L. R. P., & Vilas-Boas, E. A. (2026). Organelle-specific crosstalk between calcium and redox signaling in pancreatic β-cells in physiology and type 2 diabetes. Frontiers in Network Physiology, 6, Article 1864821. https://doi.org/10.3389/fnetp.2026.1864821

Image Credits: AI Generated

DOI: 10.3389/fnetp.2026.1864821

Keywords: pancreatic beta cells, calcium signaling, reactive oxygen species, type 2 diabetes, mitochondria, endoplasmic reticulum, NADPH oxidase, insulin secretion, mitochondria-associated membranes, oxidative stress, islet network, redox signaling

The central paradox the authors confront is chemical. ROS—including superoxide, hydrogen peroxide and hydroxyl radicals—are inherently reactive, which is precisely why uncontrolled ROS production is toxic. Yet evolution has exploited this very reactivity. Hydrogen peroxide, in particular, is small, electrically neutral and membrane-permeable, allowing it to diffuse rapidly through tissue and act as a broadcast signal. At wound margins, ROS can oxidize the cysteine residues of target proteins, reversibly altering the activity of phosphatases, kinases and transcription factors. In this way, a transient oxidative pulse functions as a second-messenger system, converting the physical fact of injury into a coordinated genetic and cellular response. The review frames this as oxidative eustress—beneficial, regulated stress—as opposed to the destructive oxidative distress associated with pathology.

What makes the new synthesis compelling is its comparative breadth. The authors assemble evidence spanning the metazoan tree, from cnidarians such as Hydra and Nematostella, which can rebuild entire bodies from small fragments, to planarian flatworms, annelids, echinoderms, fish, amphibians and mammals. In nearly every regenerative model examined, pharmacological or genetic suppression of ROS production—most often by blocking the NADPH oxidase enzymes, or NOXes, that generate superoxide at the wound—impairs regeneration. Conversely, exogenous hydrogen peroxide can, in some systems, partially rescue defective regenerative responses. The consistency of this pattern across such distantly related animals raises the tantalizing possibility that ROS signaling is not a collection of independent evolutionary accidents but a deeply rooted, possibly ancestral, component of the regeneration program.

The review is careful, however, to emphasize that conserved does not mean uniform. The regulatory roles and modes of action of ROS vary considerably among species, tissues and regenerative contexts. In some animals, ROS act primarily in the earliest wound-healing phase, recruiting immune cells and initiating proliferation; in others, oxidative signals persist into blastema formation, the mass of progenitor cells from which the new structure grows. The downstream targets differ as well: ROS have been implicated in activating Wnt, FGF, ERK and JNK signaling pathways, in modulating apoptosis, and in shaping the extracellular matrix, but the specific wiring differs from one organism to the next. This variability, the authors argue, is precisely why continued comparative study is essential. Only by mapping ROS involvement across many species can researchers distinguish genuinely common mechanisms from species-specific embellishments.

This comparative logic carries the review into deeper evolutionary waters. The evolutionary history of regeneration remains one of the great unresolved puzzles in biology. Why can a salamander regrow an entire limb while a mouse cannot? Why do some annelids regenerate their whole posterior body axis while closely related species fail at the same task? Two competing frameworks dominate the field: regeneration as an ancestral capacity repeatedly lost in lineages that traded it for other advantages, or regeneration as a trait independently gained, assembled from pre-existing wound-healing and developmental machinery. The authors argue that ROS data can inform this debate. If ROS signaling sits at the base of regenerative responses across the animal tree, it may represent an ancient, conserved layer of the process—one that lineages with poor regenerative abilities have not necessarily lost, but perhaps fail to deploy correctly.

That reframing has direct therapeutic implications, and it is here that the review makes its most forward-looking argument. Humans regenerate well in the liver, moderately in skin and bone, and hardly at all in the heart, spinal cord and limbs. If ROS signaling is a conserved trigger of regenerative programs, then modulating redox state could offer a lever to reactivate latent regenerative capacity in poorly regenerating human tissues. The concept is not fanciful. Studies in zebrafish and axolotl have shown that adjusting ROS levels can influence heart regeneration and appendage regrowth, and redox-sensitive pathways such as ERK and Wnt are already major targets of regenerative medicine research. The review suggests that a rigorous understanding of when, where and at what concentration ROS act could enable precisely timed redox interventions—boosting regeneration without tipping tissues into the oxidative damage that fuels fibrosis, inflammation and cancer.

The authors are equally candid about the obstacles. ROS are fleeting, highly reactive and difficult to measure with spatial and temporal precision, and many of the tools used to manipulate them—broad-spectrum antioxidants, generic NOX inhibitors—are blunt instruments with off-target effects. To address this, the review includes a supplementary compendium of the functional tools available to modulate ROS production in regeneration studies, cataloguing the main pharmacological inhibitors and activators as well as the principal targets for RNA interference, morpholino and knockout approaches. Standardizing and refining this toolkit, the authors contend, is a prerequisite for the field’s maturation. Without it, discrepancies between studies may reflect technical artifacts rather than genuine biological differences, muddying the comparative analyses on which evolutionary conclusions depend.

The review also identifies major knowledge gaps that will shape the next decade of work. The precise molecular targets of ROS during regeneration remain incompletely defined in most models; the dose, duration and gradients of oxidative signals at wounds are poorly quantified; and the interplay between ROS and other injury-induced signals—calcium fluxes, bioelectric gradients, mechanical forces and immune-derived cytokines—is only beginning to be explored. Another open question is whether ROS participate in the establishment of positional memory, the information that tells a regenerating blastema whether to build a hand, a tail or a head. Answering these questions will require combining live redox biosensors, genetic models and comparative approaches across species that span the breadth of regenerative ability, from master regenerators to refractory ones.

What emerges from the synthesis is a view of regeneration as a redox-gated process whose ancient logic may still be latent within our own cells. The authors, based at institutions including Université Paris Cité, Université Côte d’Azur, the University of Manchester, Sorbonne Université and École Normale Supérieure, argue that studying ROS across metazoan diversity serves a dual purpose: it illuminates how regeneration evolved, and it reveals why some species, including our own, regenerate poorly. Both lines of inquiry converge on the same practical prize—new strategies in regenerative medicine grounded in evolutionary insight rather than trial and error. If the oxidizing burst that follows injury truly is a conserved call to rebuild, then learning to read, and perhaps to amplify, that ancient signal could one day help human tissues do what salamanders have done for millions of years: grow back.

Subject of Research: The role of reactive oxygen species signaling in animal regeneration and its evolutionary and therapeutic implications

Article Title: Understanding ROS signaling and evolution during animal regeneration offers promising opportunities for novel therapeutic approaches

Article References: Vullien, A., John, J., Rampon, C., Amaya, E., Röttinger, E., & Gazave, E. (2026). Understanding ROS signaling and evolution during animal regeneration offers promising opportunities for novel therapeutic approaches. Cellular and Molecular Life Sciences. https://doi.org/10.1007/s00018-026-06392-x

Image Credits: AI Generated

DOI: 10.1007/s00018-026-06392-x

Keywords: reactive oxygen species, ROS signaling, regeneration, regenerative medicine, oxidative eustress, NADPH oxidase, evolutionary developmental biology, blastema, wound healing, redox biology, hydrogen peroxide, comparative biology