Nicotinic Signaling Stimulates Glucagon Secretion in Mouse and Human Pancreatic α-Cells

The parasympathetic nervous system regulates pancreatic islet hormone secretion via a network of intrapancreatic ganglia (1), with postganglionic fibers directly innervating the islet cells (2). Although the sympathetic system relays the severe hypoglycemic stimulus to inhibition/stimulation of insulin/glucagon release (3), the role of the parasympathetic signal is less clear (4). Acetylcholine (ACh), the principal parasympathetic neurotransmitter, is released from both enteropancreatic fibers and the neurons in the intrapancreatic ganglia (3). The alkaloid targeting ionotropic (nicotinic [nAChR]) and metabotropic receptors (muscarinic [mAChR]) within the body potentiates the release of both insulin and glucagon (5), predominantly via M3 muscarinic receptor signaling (6).

The impact of nicotinic ACh signaling on islet hormone release is less clear (7), whereas smoking has been historically viewed as a type 2 diabetes (T2D) risk factor (8), with nicotine inducing insulin resistance (9). Potential mechanisms linking nicotine consumption and the development of T2D focus on the disruption of insulin signaling, with several studies demonstrating decreased insulin sensitivity in patients with T2D who smoke (10). Dissecting the precise machinery is complicated by the broad expression of nAChR in the human body, allowing multiple potential interactions with insulin signaling. Several studies have presented diametrically opposing data on the role of nicotine in the body’s metabolism (11–14). In particular, the direct impact of nicotine on islet hormonal output is reflected by diverging views (13,15,16) that are likely to stem from a more complex cross talk between insulin and nicotine signals, involving autonomic or paracrine factors. One of the latter is glucagon, secreted by islet α-cells, which antagonizes insulin action. Elevated plasma glucagon concentrations in smokers (17) hint at a causal relationship with insulin resistance; however, little is known about the effect of exogenous nicotine on α-cell function. ACh, the natural agonist of nAChR, has been documented to induce glucagon release (18,19), mediated via α-cell mAChR (20), but involving nAChRs in chromaffin cells (21) or postganglionic (para)sympathetic neurons (22). However, these mechanisms also substantially affect the release of insulin (23), which contrasts with inconclusive reports on the nicotine effect on insulin secretion. Could nicotine have a direct effect on α-cells? Below, we explore the physiologic significance of this signal in mouse and human pancreatic islets.

All test compounds were obtained from Sigma-Aldrich (Gillingham, U.K.), apart from oxotremorine M (Insight Biotechnology, London, U.K.) and tubocurarine, CYN154806, and BIM-1 (Tocris Bioscience, Abingdon, U.K.). The choice of concentrations of the physiologic agents was guided by their systemic levels and potential vicinal availability. For instance, systemic levels of nicotine, typically ∼10 nmol/L (24), may reach 1 μmol/L after smoking (25). The synaptic concentration of ACh peaks at 260 mmol/L (26), likely resembling the peak ACh released by human islet α-cells (27); however, the expected sustained levels of the agent are lower.

The extracellular solution (EC1) contained (mmol/L) 140 NaCl, 4.6 KCl, 2.6 CaCl₂, 1.2 MgCl₂, 1 NaH₂PO₄, 5 NaHCO₃, 10 HEPES (pH 7.4; NaOH), and glucose as indicated. The pipette solution (IC1) consisted of (mmol/L) 76 K₂SO₄, 10 NaCl, 10 KCl, 1 MgCl₂, 5 HEPES (pH 7.35; KOH). The Krebs-Ringer buffer (EC2) solution for hormone secretion contained (mmol/L) 140 NaCl, 3.6 KCl, 0.5 MgSO₄, 0.5 NaH₂PO₄, 2 NaHCO₃, 5 HEPES (pH 7.4; NaOH), 1.5 CaCl₂, and glucose as indicated.

C57Bl/6 J mice (Charles River Laboratories, Harlow, U.K.) were used throughout the study, except for the experiments testing the effects of hypo/hyperglycemic culture on α-cell cholinergic signaling (Fig. 6A and B), which used mice expressing the genetically encoded Ca²⁺ indicator GCaMP3 under the control of the glucagon promoter (Glu-GCaMP3) (28). Mice were kept in a conventional vivarium with a 12-h dark/light cycle with free access to food and water. All experiments were conducted in accordance with the U.K. Animals (Scientific Procedures) Act (1986) and the University of Oxford ethical guidelines. Animals were killed by cervical dislocation, and pancreatic islets were isolated by injecting collagenase solution into the bile duct, with subsequent digestion of the connective and exocrine pancreatic tissue.

Human pancreatic islets were isolated in the Oxford Diabetes Research & Wellness Foundation Human Islet Isolation Facility according to published protocols (29,30) (Supplementary Table 1).

Unless otherwise stated, experiments were performed on acutely isolated islets or, in the case of human material, immediately after the release of the islets. For experiments aiming to emulate chronic hyperglycemia, islets were cultured for 48 h in RPMI medium supplemented with 10% FBS, 100 international units/mL penicillin, 100 μg/mL streptomycin (Life Technologies, Paisley, U.K.), and glucose, as indicated, in the atmosphere of 5% CO₂, absolute humidity, +37°C. Recombinant sensors (Downward Green DAG [diacylglycerol], Upward Red cADDis) were delivered via BacMam (Montana Molecular, Bozeman, MT) vectors at 10⁵ international units per islet, followed by 36-h culturing (as above) to express the proteins.

Time-lapse imaging of intracellular Ca concentration ([Ca²⁺]_i) in freshly isolated mouse islets was performed on an inverted Zeiss AxioVert 200 microscope equipped with Zeiss 510-META laser confocal scanning system. Islets were preloaded with 6 μmol/L Fluo-4 (Molecular Probes) at room temperature for 90 min. The dye was excited at 488 nm, and emission was collected at 530 nm using a 512 × 512 frame scan (pixel dwell 6 μs; imaging frequency 0.25 Hz). Imaging of [Ca²⁺]_i in groups of mouse islets was performed on an Axiozoom.V16 microscope using GCaMP3 that was engineered to express selectively in islet α-cells. GCaMP3 was excited at 470 nm, with the emission collected at 530 nm at an acquisition interval of 10 s.

For [DAG]_i imaging, isolated islets were dispersed into a cell suspension, plated on glass coverslips, and coinfected with Downward Green DAG and Upward Red cADDis cAMP probes. The two signals were subsequently imaged from the cells using a Zeiss AxioZoom.V16 zoom microscope, using the excitation of 470 nm (550 nm) and emission of 530 nm (585 nm) for the DAG (cAMP) sensors. Images were taken every 30 s. α-Cells were identified within islets by their positive response to glutamate (31) or adrenaline, which selectively increases cAMP and [Ca²⁺]_i in α-cells (32). The EC1 bath solution at 37°C was perifused at 60 μL × min⁻¹ continuously throughout the imaging experiments.

Electrophysiologic measurements were performed in α-cells within intact islets of Langerhans using a perforated-patch whole-cell configuration. The borosilicate glass pipette tip was dipped into IC1 and back-filled with the same solution containing 0.16 mg/mL amphotericin B (tip resistance 4–6 MΩ). After perforation of the plasma membrane, membrane potential was recorded in current-clamp mode, with zero current injection, using an EPC10 amplifier controlled by Pulse software (HEKA Elektronik, Lambrecht, Germany). α-Cells were identified by functional fingerprinting (33). The EC1 bath solution, supplemented with glucose or other agents as indicated, was maintained at +34°C and perifused continuously at 200 μL × min⁻¹.

After cervical dislocation, the mouse aorta was cannulated by ligating 1) above the celiac artery and 2) below the superior mesenteric artery. The pancreas was perfused with EC2 solution at 0.45 mL × min⁻¹. The perfusate was maintained at +37°C with a tube heater (Warner Instruments, Hamden, CT) and a heated rodent operating table (Harvard Apparatus, Holliston, MA). The effluent was collected via a portal vein cannula using a Teledyne (Thousand Oaks, CA) ISCO Foxy R1 fraction collector. The pancreas was initially perfused for 20 min with 1 mmol/L glucose before commencing the experiment to establish the basal rate of secretion.

Batches of 10–20 size-matched freshly isolated mouse islets were preincubated in 1 mL EC2 solution containing 1 mmol/L glucose and 0.2% BSA for 30 min at +37°C, followed by a 1-h test incubation in 1 mL of the same medium supplemented as indicated. Glucagon was determined with RB306 (Svar Life Science/Eurodiagnostica, Malmö, Sweden) radioimmunoassay.

Image sequences were analyzed using open-source FIJI software (https://fiji.sc/Fiji). The numeric data were analyzed using IgorPro (WaveMetrics, Portland, OR). To calculate partial areas under the curve, the recording was split into 30-s intervals, and the area under the curve was computed for each interval using trapezoidal integration (31).

Public domain diabetes and smoking prevalence data available for parliamentary constituencies and National Health Service wards, respectively, were recomputed per individual U.K. postcode. The map (Fig. 6C) represents a color-coded metric reflecting correlation between the two indices, computed as $(d−d¯)(s−s¯)/σd/σs,$ where $d(s), d¯(s¯), σd(σs)$ are the per-postcode local nationwide average and SD of the prevalence of diabetes (smoking).

The sample sizes are specified in the figure legends. The experiments on human islets were performed on islets isolated from five donors. A minimum of five mice were used for each condition, with n taken as the number of cells (electrophysiology, imaging) or groups of islets (other studies). Statistical analysis was performed using R. Data are presented as mean ± SEM. The Mann-Whitney (Wilcoxon) test was used to compute the significance of differences between independent (dependent) samples. Multiple comparisons were performed using the Kruskal-Wallis (Friedman) test with Nemenyi post hoc analysis for independent (dependent) samples.

Data are available on request.

At 3 mmol/L glucose, mouse and human α-cells displayed spontaneous [Ca²⁺]_i oscillations, whereas [Ca²⁺]_i was stable in β-cells (Fig. 1A and B). ACh (10 μmol/L) elicited a reversible elevation of [Ca²⁺]_i in both cell types (Fig. 1A–C), but the response was greater and more sustained in α-cells (Supplementary Movies 1 and 2).

The contribution of metabotropic or ionotropic signaling to the ACh-stimulated increase in [Ca²⁺]_i in islet cells (Fig. 1A–C) was dissected using specific agonists of mAChR (oxotremorine M) and nAChR (nicotine). Although oxotremorine M (10 μmol/L) rapidly increased [Ca²⁺]_i in both cell types (Fig. 1C), nicotine (10 μmol/L) had no effect in β-cells but induced a significant increase in [Ca²⁺]_i in 71% of α-cells (Fig. 1C). The α-cell selectivity of nicotine action (Fig. 2A) was conserved in human islets (Fig. 2B), although the magnitude of the stimulation and the fraction of the responsive α-cells (37%) were lower than in the mouse counterparts. The onset time of the half-maximal effect of nicotine on α-cell [Ca²⁺]_i (40 ± 2 s) was similar to that of ACh (38 ± 2 s), whereas the response to oxotremorine was significantly delayed (55 ± 2 s) (Fig. 2C). The stimulatory effect of nicotine on [Ca²⁺]_i dynamics in α-cells was preserved at high (20 mmol/L) glucose (Fig. 2A) and was well reproducible using GTS-21 or succinylcholine, selective agonists of α7 and α1 nAChR subunits, respectively (Fig. 2A).

We discounted the role of intraislet sympathetic nerve terminals in the nicotine-dependent stimulation because the effect persisted in the presence of β-adrenergic blocker propranolol and after a 24-h in vitro culture of islets, which induces a significant degradation of most intraislet neurons (34) (Fig. 2A).

The [Ca²⁺]_i responses to cholinergic stimulation in α-cells (Fig. 2A) matched the effects of ACh receptor agonists on glucagon secretion from isolated islets (Fig. 2E). The nonselective ACh receptor agonist carbachol (used in the static incubation experiments because it cannot be inactivated by ubiquitous cholinesterase), as well as oxotremorine M, nicotine, GTS-21, and succinylcholine, stimulated glucagon release from isolated mouse islets at 3 mmol/L glucose by 3.3 ± 0.4, 4.3 ± 0.8, 2.5 ± 0.3, 2.2 ± 0.3, and 2.2 ± 0.6-fold, respectively. The stimulatory effect of nicotine on glucagon secretion was preserved at high (20 mmol/L) glucose (Fig. 2E).

We estimated the contribution of the two ACh signaling components in islet α-cells using inhibitors of mAChR (atropine) and nAChR (tubocurarine, α-bungarotoxin). Atropine (10 μmol/L) significantly reduced ACh-stimulated α-cell [Ca²⁺]_i dynamics or carbachol-enhanced glucagon secretion from pancreatic islets (Fig. 2D and F). Tubocurarine (5 μmol/L) or α-bungarotoxin (100 nmol/L) led to a mild but significant reduction (19.8 ± 9.3% and 37.2 ± 8.7%, respectively) of the α-cell Ca²⁺ response to ACh (Fig. 2D). Tubocurarine, however, had only a limited impact on carbachol-stimulated glucagon secretion (Fig. 2F) and did not affect the ACh-induced [Ca²⁺]_i elevation in β-cells (Supplementary Fig. 1C).

We explored the intracellular pathways mediating the cholinergic stimulation in α-cells, focusing on hallmarks of metabotropic (elevation of intracellular [DAG]) or ionotropic (depolarization of plasma membrane) signaling. Added at 3 mmol/L glucose, ACh (10 μmol/L) significantly and reversibly increased [DAG]_i in 69% of α- and 94% of β-cells (Fig. 3A). On average, fluorescence of the DAG sensor increased by 9 ± 5% and 23 ± 7% in α-and β-cells, respectively (Fig. 3B).

Phospholipase C–mediated G_q signaling progresses via DAG-dependent activation of protein kinase C (PKC) and IP₃-induced release of Ca²⁺ into the cytosol from luminal stores. The removal of extracellular Ca²⁺ (with 5 mmol/L EGTA) did not affect ACh-induced elevation of [Ca²⁺]_i in β-cells or a majority of α-cells (Supplementary Fig. 1I). However, a sizeable (36%) fraction of α-cells did not respond to ACh stimulus in the absence of extracellular Ca²⁺ (Fig. 3C and D).

Assayed using the perforated-patch configuration that maintains intracellular metabolism intact, α-cell electrical activity exhibited divergent responses to ACh and nicotine. Whereas the acute application of 10 μmol/L ACh triggered a transient repolarization (Fig. 3E arrow) and cessation of action potential firing in α-cells (70% of cells), nicotine induced a long-lasting membrane depolarization and intensified α-cell electrical activity (75% of cells) (Fig. 3E and F).

The heterogeneity of α-cell response to ACh in the absence of extracellular Ca²⁺ (Fig. 3C) indicated that the ACh effect, in part, was mediated by Ca²⁺ influx through plasmalemmal voltage-gated Ca²⁺ (Ca_V) channels. To identify the sources of the Ca_V conductance involved in the ACh-induced elevation of [Ca²⁺]_i and glucagon secretion in α-cells, specific inhibitors of l-type (isradipine) and P/Q-type (ω-agatoxin) Ca_V channels were used. Isradipine (2 μmol/L) or ω-agatoxin (200 nmol/L) significantly attenuated the carbachol-stimulated glucagon secretion by 55 ± 8% and 84 ± 5%, respectively (Fig. 4A). Neither antagonist, however, cancelled the stimulatory cholinergic effect completely. ω-Agatoxin significantly attenuated the increase in α-cells [Ca²⁺]_i induced by ACh by 31 ± 5%, whereas the effect of isradipine was not significant (Fig. 4B). Nicotine-induced [Ca²⁺]_i increases in α-cells were, similarly, reduced by ω-agatoxin and isradipine (by 52 ± 8% and 63 ± 9%, respectively).

Because nicotine was demonstrated to activate the TRPA1 cation channel in other systems (35), we assessed a possible contribution of this component in α-cells by specifically inhibiting the channel with AP18 (10 μmol/L) in the presence of nicotine (Fig. 4B). The nonsignificant effect of AP18 was contrasted by mild stimulation of TRPA1 with J-010 (20 μmol/L) (Fig. 4B).

The release of luminal Ca²⁺, which contributes to the elevation of [Ca²⁺]_i induced by ACh stimulus, can be accomplished via GPCR/G_q/IP₃ machinery, used by mAChR, or Ca²⁺-induced Ca²⁺ release, initiated by nAChR (36). Having depleted the endoplasmic reticulum (ER) Ca²⁺ stores by preincubating islets in the SERCA-Ca²⁺ ATPase inhibitor CPA (10 μmol/L) for 20 min, we observed that the ACh-induced [Ca²⁺]_i increase in α-cells was substantially reduced (Fig. 4C). The response to the mAChR agonist oxotremorine M was likewise attenuated by ER Ca²⁺ depletion (−52 ± 12%). However, the effect of nicotine on α-cell [Ca²⁺]_i was unaffected in islets pretreated with CPA (Fig. 4C). Notably, a specific inhibitor of the DAG target, PKC (BIM-1, 200 nmol/L), had no effect on the elevation of [Ca²⁺]_i induced by ACh (Fig. 4C).

Islet cell hormonal response to global stimuli is modulated by several acute local factors released by the neighboring cells (37). The impact of the local environment on the cholinergic stimulation of islet α- and β-cells was assessed by in situ perfusing the mouse pancreas. In the presence of 1 mmol/L glucose, nicotine significantly increased the secretion of glucagon (from 53.7 ± 5.4 to 129.3 ± 8.3 pg/min; P < 0.05 vs. basal) (Fig. 5A). Notably, the application of nicotine reduced subsequent glucagon secretion at high (10 mmol/L) glucose (from 9.3 ± 0.5 to 5.4 ± 0.7 pg/min; P < 0.05) (Fig. 5A and B).

In rodent models, cholinergic signaling reportedly attenuated the release of the inhibitory hormone somatostatin, secreted by islet δ-cells (38). We probed the impact of somatostatin on ACh signaling in α-cells using a somatostatin receptor (SstR) antagonist, CYN 154806, that at 100 nmol/L selectively inhibits the receptor isoform expressed in α-cells, SstR₂ (39). At 3 mmol/L glucose, acute application of CYN154806 stimulated basal [Ca²⁺]_i dynamics in 58% of α-cells (Fig. 5C and D). ACh exerted a further significant stimulatory effect in the presence of CYN154806. Notably, the magnitude of the ACh effect was not affected by CYN154806 (Fig. 5D).

We tested the impact of chronic hyperglycemia on the cholinergic signaling in islets precultured at different concentrations of glucose (5, 11, or 30 mmol/L) for 72 h. The effect of ACh on [Ca²⁺]_i dynamics was imaged in the three groups simultaneously (Fig. 6A). Basal α-cell [Ca²⁺]_i dynamics was significantly higher in islets precultured at 30 mmol/L glucose (Fig. 6C). Nicotine (10 μmol/L) or ACh (10 μmol/L) induced a rapid and reversible increase in [Ca²⁺]_i in all groups, but the effect was significantly higher in islets cultured at 30 mmol/L glucose compared with islets cultured at 5.5 or 11 mmol/L glucose (Fig. 6C), suggestive of the upregulation of this signaling under chronic glucotoxicity. These observations were supported by data on glucagon secretion in isolated islets (Fig. 6D) that demonstrated an enhanced response to nicotinic and especially cholinergic stimulation in islets precultured at 30 mmol/L glucose.

Finally, we probed the effect of chronic (72-h) exposure to nicotine or carbachol on the acute responses to these stimuli. Preculturing the islets to nicotine significantly attenuated the responses to both compounds, whereas chronically administered carbachol enhanced the sensitivity of islet α-cells to nicotine, at the level of [Ca²⁺]_i dynamics (Fig. 6C).

The availability of nicotinic ACh signaling in α-cells would make glucagon secretion sensitive to the systemic levels of alkaloid nicotine, which is conventionally consumed in various forms by 22.3% of the global population (https://www.who.int/news-room/fact-sheets/detail/tobacco). Computed per postcode area within England and Wales, the percentage of smokers (Supplementary Fig. 2B) displayed a moderate correlation (Pearson r = 0.44; P < 0.05) with diabetes prevalence (Fig. 7A and Supplementary Fig. 2B).

We report differential effects of ACh signaling in cellular subpopulations of mouse and human pancreatic islets. ACh signal promoted the production of DAG in islet β-cells, whereas in α-cells, it exhibited a potent nicotinic component, contributing, alongside its muscarinic counterpart, to the release of glucagon. Activation of nAChR had a small depolarizing effect in α-cells, leading to increased action potential firing and stimulation of Ca²⁺ influx via P/Q and L-type voltage-gated Ca²⁺ channels (Fig. 4B), thereby inducing the secretory response.

Originating from Glut2-expressing brainstem and hypothalamic neurons that stimulate the vagal signal, the parasympathetic response to hypoglycemia is subsequently relayed to ACh release from intraislet terminals (2,40). Parasympathetic signaling is believed to sense small physiologically relevant drops in glucose in the range between 4.2 and 4.7 mmol/L (22). In humans, α-cells were reported to release ACh in a glucose-dependent manner, thereby providing an alternative (arguably, predominant) source of the agent in the islet vicinity.

Of two types of ACh receptors (Fig. 7B), mAChR is believed to be most abundant in the islets; however, mRNA of nAChR subunits is detectable in all types of islet cells (41). The predominance of Chrna1 and Chrna7 isoforms (42) suggests that α-cells are likely to express a mixture of neuronal- and muscle-type nAChRs that have varying affinity (43,44) and response kinetics (45) to nicotine. In line with this, both GTS-21 and succinylcholine significantly enhanced [Ca²⁺]_i dynamics and glucagon secretion in islet α-cells (Fig. 2A and E). In particular, the glucagonotropic effect of GTS-21 at the whole-body level (46) may, therefore, have a direct modality.

The nAChR is a ligand-gated cation channel inducing the depolarization of plasma membrane with subsequent Ca²⁺ entry via Ca_V channels. Pharmacological inhibition of two Ca_V channel types significantly reduced the acute α-cell nicotinic response at the level of [Ca²⁺]_i (Fig. 4B) and, more profoundly, hormone secretion (Fig. 4A). We attribute this difference to the depletion of the luminal Ca²⁺ stores during the prolonged incubations used for the secretion studies.

Our findings demonstrate that both nAChR and mAChR are involved in the cholinergic stimulation of the α-cell (Fig. 2C and D), exerting additive effects on glucagon secretion. Different kinetics suggest distinct roles for the two types of receptors in the control of α-cell function. Given its faster onset, the nicotinic signaling could either operate as a fine-tuning mechanism or, vice versa, provide a kick start for the larger and more sustained increases in [Ca²⁺]_i, induced via the mAChR that was reported to function in an autocrine manner, in human α-cells (27). Thus, the smaller nicotinic stimuli inducing a specific response in α-cells would likely go unnoticed by β-cells. In line with previous reports (7,23), the overall long-term secretory output, which depends on the amplification of the initiating Ca²⁺ signal (47), would be higher in β-cells that exhibited a significant elevation of the PKC agonist, DAG (Fig. 3).

The stimulatory effect of nicotine seemed to be limited to a specific subset of α-cells (71%), complemented by [DAG]_i imaging data, demonstrating a population of α-cells not responsive to ACh (Fig. 3A and B). Could there be separate subsets of α-cells favoring nAChR and mAChR stimuli? Our data do not rule out the exclusive nAChR/mAChR specialism of islet α-cells, which, however, cannot be tested reliably in this experimental system, because the elevation of [Ca²⁺]_i is implicated in both mechanisms.

The heterogeneity in the α-cell [DAG]_i response may originate from variable activation of nicotinic/muscarinic pathways across the α-cell population. The smaller magnitude of the [DAG]_i response compared with β-cells supports the notion of prominent nicotinic signaling in a large fraction of α-cells. Interestingly, the small fraction of non–α-cells (6%; identified by the negative cAMP response to adrenaline) that were not responsive to ACh may represent δ-cells. With adrenaline responsiveness similar to that of β-cells, δ-cells reportedly do not increase [DAG]_i on stimulation by ACh (38). The latter had a low impact on somatostatin secretion from the human pancreas (48), in line with the somatostatin independence of the cholinergic effect on α-cell [Ca²⁺]_i dynamics (Fig. 5).

Human islets displayed a potent response to ACh, with á-cells manifesting a strong nicotinic sensitivity (Figs. 1B and 2B). Although the physiologic effects of nicotinic signaling on glucagon secretion are important, they likely take even greater precedence when applied to the pathophysiologic effects of exogenous nicotine. Smoking causes a fall in insulin sensitivity in both control participants and patients with T2D, and our findings prompt a possible mechanism for this phenomenon. Selective stimulation of α-cells by nicotine, reported here, would increase glucagon secretion and hepatic glucose output, thereby counteracting the effect of insulin. In T2D, this stimulus would compound hyperglycemia in destabilizing glycemic control, which agrees well with the reports of higher HbA_1c levels in smokers with T2D (10). Enhanced α-cell sensitivity to nicotinic/cholinergic stimulus (Fig. 6C and D) may exacerbate this effect. On the other hand, chronic sustained hyperglucagonemia is likely to impair glucose counterregulation, increasing the risk of hypoglycemic episodes during insulin therapy, potentially via glucagon or nicotine receptor desensitization.

Although vagal signaling plays a definitive role in the glucose-induced enhancement of islet blood flow, there are opposing accounts regarding the scale of direct parasympathetic innervation of human islet cells (2,49). The contribution of the spillover of ACh from the intraislet vessels is likely to be insignificant because of the abundance of acetylcholinesterase (50). Although autonomic ACh seems to have an unclear influence on islet hormonal output in humans, reports of local ACh production by α-cells (27) may suggest a different role for the nicotinic component of α-cell cholinergic signaling. In that context, α-cells could use the fast nicotinic machinery to detect small or transient extracellular ACh signals and amplify them by releasing their endogenous ACh that would target neighboring β- or δ-cells (7).

In conclusion, we report the nicotinic sensitivity of the potent machinery linking parasympathetic signaling to glucagon secretion in mice and humans. Because overall cholinergic stimulation of pancreatic hormone release is blunted in T1D and advanced T2D, because of antecedent iatrogenic hypoglycemia (22), the role of the nicotinic signal in the gradual loss of glucose sensitivity represents a promising goal for future endeavors. Likewise, the therapeutic use of the nicotinic component of islet ACh signaling holds promise for the correction of defective counterregulation in T2D. The significant association between the percentage of smokers and diabetes prevalence (Fig. 6C) may reflect the link to common pathophysiologic traits or socioeconomic factors. Speculatively, however, the causal association between the two indices could involve nicotine-induced hyperglucagonemia that would dynamically decrease insulin sensitivity of peripheral tissues and elevate basal glucose.

Funding. A.H. was a recipient of a Diabetes UK PhD Studentship. During the initial stages of the project, A.I.T. held an Oxford Biomedical Research Council postdoctoral fellowship. Q.Z. is a Diabetes UK RD Lawrence Fellow.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.H., A.S., J.G.K., R.G., T.G.H., and M.D. were responsible for methodology and investigation, data curation, and visualisation. Q.Z. and P.R. were responsible for conceptualization, supervision, writing, reviewing, and editing. P.R.V.J. was responsible for conceptualization, supervision, and methodology. A.I.T. was responsible for data curation, investigation, visualization, conceptualization, writing, reviewing, and editing. A.I.T. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.