LRRC8 proteins in the pancreas and normal islet morphology upon β-cell-specific Lrrc8a disruption. a Western blot detection of the five VRAC subunits LRRC8A, -B, -C, -D, and –E in lysates of purified islets of Langerhans (left lanes) or total pancreas (right) from wild-type mice. β-actin, loading control. Arrowheads highlight specific bands as determined by previous knock-out controls. b Immunofluorescent detection (green) of LRRC8D in pancreatic sections from knock-in mice expressing a LRRC8D-tdTomato fusion protein, co-stained (in red) for insulin (above) or glucagon (below). Left panels, individual channels; right panels, overlays, with co-localization yielding yellow. Note that β-cells express much more LRRC8D than the surrounding tissue. c Western blot of lysates from total pancreas or purified islets probed for LRRC8A, insulin, Kir6.2 (K_ATP channel subunit) from Lrrc8a^lox/lox mice lacking Cre expression (−) or expressing Cre specifically in β-cells (+). β-actin, loading control. Right bottom, quantification of LRRC8A expression in pancreatic islets normalized to Lrrc8^lox/lox control. N = 3 independent experiments. Error bars, mean ± SEM. **P < 0.01 (Student’s t-test). d Lrrc8a promoter-driven β-gal expression (X-gal staining, blue dots) in islets. Dotted lines highlight islets of Langerhans, insets higher magnification of boxed area. Cells co-stained with eosin Y (pink). e Hematoxylin/eosin (H&E) stained formalin-fixed pancreatic sections of control (Lrrc8a^lox/lox) or βc-Δ8a mice (lacking LRRC8A in β-cells). f Immunofluorescent staining of pancreatic islets of Lrrc8a^lox/lox and βc-Δ8a animals for insulin (green) and glucagon (red), respectively. g Insulin levels of whole pancreas lysates normalized to control (Lrrc8a^lox/lox). h Pancreatic mass of Lrrc8a^lox/lox and βc-Δ8a animals determined as ratio of organ to body weight. i β-cell mass as defined as ratio of insulin-positive cell area to complete pancreatic tissue area, multiplied by pancreatic weight. j Body weight of Lrrc8a^lox/lox (black bars) and βc-Δ8a animals (white bars) at 3, 10 and 30 weeks of age. g–j Mean values ± SEM are shown. Differences between genotypes are not significant in unpaired t-test: g p = 0.185, n = 7; h p = 0.609, n = 3; i p = 0.968, n = 6. j p = 0.086 at 3 weeks, p = 0.05 at 10 weeks, and p = 0.214 at 30 weeks, n = 16 (Lrrc8a^lox/lox), n = 11 (βc-Δ8a). Scale bars in b: 10 µm, in d–f: 50 µm

Hypotonicity-induced cell swelling and VRAC currents in mouse β-cells. a Cell volume as monitored in individual β-cells using calcein fluorescence. The extracellular isotonic solution was replaced by a 30% hypotonic solution (210 mOsm) during the time indicated by the bar. Averaged curves ± SEM (dotted lines), n = 16 and n = 18 for Lrrc8a^lox/lox and βc-Δ8a β-cells, respectively. Data are representative of three independent experiments. b Representative time course of I_Cl,vol of an isolated primary β-cell provoked by superfusion with hypotonic solution (240 mOsm) without or with 20 µM DCPIB as indicated. Currents elicited at −80 to 80 mV by voltage ramps (see c) are plotted. c Typical current-voltage (I–V) curves of hypotonicity-provoked I_Cl,vol elicited by the ramp protocol shown below, from indicated genotypes and conditions. d Representative current traces obtained with the voltage step protocol shown at bottom. The upper traces show control (Lrrc8^lox/lox) currents after full activation of I_Cl,vol, taken at the time indicated by arrow in b. Even at +100 mV, β-cell I_Cl,vol only slightly inactivates. The lower traces are from a βc-Δ8a β-cell exposed for 15 min to hypotonicity. e I_Cl,vol current densities at −80 mV of Lrrc8a^lox/lox and βc-Δ8a β-cells under iso- and hypotonic conditions. Mean values ± SEM; ***p < 0.0005 (one-way ANOVA, Tukey’s test); number of cells given in bars; iso: isotonic solution (310 mOsm), hypo: hypotonic solution (240 mOsm). f Ion selectivity of β-cell I_Cl,vol. Example traces (from the same cell) obtained with the ramp protocol after maximal I_Cl,vol activation, in the presence of extracellular chloride (black trace) or iodide (green trace). The shifts of reversal potentials E_rev with I⁻ to more negative values indicated a permeability ratio pI⁻/pCl⁻ = 1.44 ± 0.06 (n = 6)

Cell swelling and induction of LRRC8/VRAC currents in β-cells by high extracellular glucose. a Cell volume as monitored in individual β-cells loaded with calcein. Cell bath was switched from a 3 mM to a 25 mM glucose-containing isotonic solution at the time indicated by the arrow. Shown are mean values ± SEM (dotted lines), n = 15 and n = 23 for Lrrc8a^lox/lox and βc-Δ8a β-cells, respectively, and are representative of three independent experiments. Green trace corresponds to bath change without changing the glucose concentration, a control to exclude perfusion artefacts (n = 6). b Time course of anion current activation by superfusion with 20 mM glucose in isotonic saline (310 mOsm). Application of 20 µM DCPIB blocked the currents. Minimal and maximal currents elicited at −80 or 80 mV by voltage ramps (as in Fig. ) are plotted. c Representative current traces obtained after >10 min superfusion with 20 mM glucose from the indicated genotypes. Currents were elicited by the voltage protocol shown in Fig. . The upper traces show currents at the time indicated by the arrow in b. d I_Cl,vol current densities at −80 mV of Lrrc8a^lox/lox and βc-Δ8a β-cells under indicated conditions. Mean currents ± SEM; *p < 0.05; ***p < 0.0005 (one-way ANOVA, Tukey’s test); number of cells indicated in bars

Hypotonicity-induced depolarization of β-cells. Gramicidin-perforated patch-clamp recordings of primary mouse β-cells. a Resting membrane potential of control Lrrc8a^lox/lox or βc-Δ8a β-cells at 1 or 5 mM glucose concentrations. Mean values ± SEM indicated by error bars. b Representative trace for hypotonicity-induced depolarization of the membrane potential V_m in control Lrrc8a^lox/lox animals at 5 mM glucose. TOLB, tolbutamide. c ΔV_m after applying hypotonicity, or following washout with isotonic solution of Lrrc8a^lox/lox or βc-Δ8a cells at 5 mM glucose, calculated from data as indicated in b. Spiking cells (see g and h) were excluded from analysis. d Representative voltage trace showing lack of hypotonicity-induced depolarization of βc-Δ8a β-cells at 5 mM glucose. e ΔV_m of Lrrc8a^lox/lox or βc-Δ8a cells as in c, but at 1 mM glucose. f Representative trace for Lrrc8a^lox/lox β-cells challenged with hypotonicity in the presence of 1 mM glucose. g Percentage of cells responding with electrical activation following hypotonic stimulation. h Voltage trace of an Lrrc8a^lox/lox β-cell responding with electrical activity to hypotonicity in the presence of 5 mM glucose. *p < 0.05, **p < 0.01 (one-way ANOVA, Tukey’s test)

Glucose-induced electrical activity and intracellular Ca²⁺ response in Lrrc8a^lox/lox and βc-Δ8a cells. a, b Representative voltage traces for glucose- and tolbutamide-induced electrical activity of control Lrrc8a^lox/lox (a) and βc-Δ8a (b) β-cells obtained with gramicidin-perforated patches in the current clamp mode. Enlarged traces are shown in red boxes, with corresponding time periods indicated by boxes in the complete trace at left. c Percentage of electrically active cells plotted against time after glucose application. The response of βc-Δ8a cells seemed delayed by about 100 s (dotted lines), but the difference between genotypes failed to be statistically significant with 24 control and 18 βc-Δ8a β-cells. d Mean frequency and amplitude of spikes during bursts of action potentials elicited by 15 mM glucose. e Mean frequency and amplitude of spikes elicited by 300 µM tolbutamide. f Individual traces for Fura-2 fluorescence ratios elicited by excitation at λ = 340 and 380 nm (indicative of [Ca²⁺]_i). 15 mM glucose was added at t = 60 s. g Mean ratio of Fura-2 fluorescence over time. Time points at which differences between the two genotypes reached statistical significance are indicated. *p < 0.05, ^#p < 0.01 (two-way ANOVA, Bonferroni multiple comparisons); mean values ± SEM, 69 control and 71 βc-Δ8a β-cells. h Cumulative plot of cells that have reached peak levels of [Ca²⁺]_i after addition of 15 mM glucose as function of time. The difference is statistically significant (p = 0.005, Kolmogorov-Smirnov test). i Peak value of calcium response of β-cells of both genotypes. The difference is not significant. Number of cells indicated in columns

Effect of LRRC8A/VRAC disruption on insulin secretion in vitro and in vivo and on glucose tolerance. a Insulin release from single islets stimulated by high glucose, tolbutamide, and potassium. Islets were incubated for 30 min under the indicated conditions and insulin concentrations in the supernatant were measured. Number of islets is indicated in bars (derived from 5 mice per genotype (25 mM glc), 1 mouse per genotype (TOLB), 3 mice per genotype (KCl), 2 mice per genotype (TOLB + glc)). b Insulin release of single islets under the indicated conditions during the first 8 min of incubation. Number of islets (from three mice per genotype and condition) is indicated in bars. ***p < 0.001 (one-way ANOVA, Tukey’s test). c Glucose tolerance test. Changes in blood glucose levels following intraperitoneal injection of glucose (2 mg/g body weight) in Lrrc8a^lox/lox, Lrrc8a^+/+;Ins-Cre⁺ and βc-Δ8a animals was followed over time. d Insulin tolerance test. Changes in blood glucose levels over time following intraperitoneal injection of insulin (0.75 U/kg body weight). c, d Number of mice (male adults, 10–15 weeks old) for each experiment is indicated. *p < 0.05, **p < 0.01 (two-way ANOVA, Bonferroni multiple comparisons); mean values ± SEM. e Glucose-stimulated insulin secretion in vivo. Changes in the plasma insulin concentration after intraperitoneal injection of a glucose load (2 mg/g body weight) over time. Number of mice (adult males and females, 10–15 weeks of age) is indicated. Unpaired t-test for insulin levels at 2 min after glucose injection yields p = 0.07