SUPPLEMENTARY MATERIALS

RESEARCH DESIGN AND METHODS

Measurement of [Ca2+]i.Dispersed  cells obtained by trituration of islets with a pipette were seeded on confocal dishes and cultured in the culture media. Approximately 90% of stably adhered cells showed insulin positive staining.The  cells were identified on the basis of their large diameter and granular appearance. Adhered cells were washed with KRB buffer containing 0.1% BSA and loaded with 1 mol/l Fluo-3 AM (Molecular Probe, Eugene, OR) at 37°C for 30 min. After washing with KRB buffer containing 0.1% BSA, an appropriate amount of glucose was supplemented. In some cases, Fluo-3 AM was loaded in the presence of 12 mmol/l glucose in KRB buffer and 0.1% BSA. Changes in fluorescence in  cells were determined as described previously (1).

Measurement of intracellular cADPR concentration. Level of cADPR was measured using a cyclic enzymatic assay as described previously(2). For the experiment, Neurospora crassa NADase and Aplysia ADPR-cyclase were purified according to the methods described by Cho et al. (3) and Lee et al. (4), respectively.

Measurement of intracellular NAADP concentration. Level of NAADP was measured using a cyclic enzymatic assay as described previously (2).

Separation of organelles.Membrane, lysosome, and insulin-containing secretory granules were separated as described (31). Briefly, islets isolated from 25 mice were homogenized with 23G syringe in 5 mmol/l HEPES (pH 7.4) containing 10 g/ml leupeptin, 2 g/ml aprotinin, and 0.25 mol/l sucrose. After centrifugation of the homogenate at 3000 x g for 10 min, supernatants were collected and loaded onto a sucrose density gradient (0.4–2 mol/l). After centrifugation at 110,000 x gfor 18 h, 17 fractions were collected from the bottom of the tube. The distribution of the organelles was assessed by determining Na+/K+ ATPase level, -D-galactosidase activity or insulin

Reconstitution study. Islets preincubated with 12 mmol/l glucose in KRB buffer at 37oC for 40 min were stimulated with GLP-1 or PKA- and Epac-specific activator at 37°C for 30 s, and the samples were quenched in an ice-water bath and lysed by sonication. The ice-cold lysates were mixed with each fractionobtained by sucrose gradient and then incubated at 37°C for 2 min. Reaction was stopped by adding perchloric acid and used to determine the levels of NAADP and cADPR as described above.

Analysis of CD38 in organelles.Each fraction was lysed by adding ice-cold lysis buffer (1% Triton X-100, 10% glycerol, 100mmol/l NaCl, protease inhibitors, and 20 mmol/l HEPES, pH 7.2). The lysed fractions were incubated overnight withanti-mouse CD38 or anti-mouse IgG Abs at 4oC, and then further incubated with protein G-agarose at 4oC for 5 h.NADase activity in the immunoprecipitates and supernatants was determined by using 1,N6-etheno-NAD (-NAD) as described (6).

Measurement of insulin.Isolated isletswere cultured overnight. For insulin secretion, each tube containing 5 islets was analyzed. The analysis was performed 5 times for each sample in order to minimize the errors. Cells were rinsed twice with KRB buffer and pre-incubated in 12 mmol/l glucose containing KRB buffer for 40 min at 37°C. GPN, bafilomycin A1 or vehicle was applied at 37°C for 20 min. After aspiration and washing with 12 mmol/l glucose KRB buffer containing each blocker, thecells were stimulated with 1 ml of 12 mmol/l glucose KRB buffer containing either GPN, bafilomycin A1 or vehicle, supplemented with differentstimuli for 30 min at 37°C. The medium containing secreted insulin was collected and assayedby using the radioimmunoassay according to the manufacturer’sinstructions.

SUPPLEMENTARY FIGURE LEGENDS

Supplementary Figure 1.The two cAMP-sensitive molecules, PKA and Epac, play a role in GLP-1-induced Ca2+ signal.A: GLP-1-induced Ca2+ increase. B-C: The effect of GLP-1 on Ca2+ signal in the presence of PKA inhibitors. D: The GLP-1-induced insulin secretion was reduced but not completely blocked by treatment with H89 or Rp-cAMP. *P < 0.005 versus 2.8 mmol/l glucose. **P < 0.005 versus 12 mmol/l glucose. #P < 0.01 versus 12 mmol/l glucose + 10 nmol/l GLP-1. E-F: Activators of PKA (100 mol/l 6-Bnz-cAMP) and Epac (100 mol/l 8-pCPT-2'-O-Me-cAMP) induced a rapid and sustained increases in [Ca2+]i. G-H: PKA activator-induced Ca2+ augmentation was blocked by pretreatment of the cells with PKA inhibitor, but not Epac activator-induced Ca2+ increase.

Supplementary Figure 2. The acetylcholine-induced Ca2+ signaling was inhibited by the IP3 receptor blocker, xestospongin C. A:  cells preincubated with 12 mmol/l glucose were stimulated by 100 mol/l Ach.  cells pretreated with 2 mol/l xestospongin C were stimulated by Ach in the extracellular calcium positive (B) or negative (C) condition.

Supplementary Figure 3.Mechanism of cADPR transport into  cells. Transport experiments were executed by a rapid oil-stop procedure previously established in our laboratory (7) and cADPR transported was measured using a cyclic enzymatic assay as described in the manuscript. Pancreatic islets preincubated with 12 mmol/l glucose were treated with 100 mol/l cADPR. cADPR transport in  cells is Ca2+-dependentand partially blocked by dipyridamole, an inhibitor of equilibrative nucleoside transporters. *P < 0.01 versus control, #P < 0.01 versus 100 mol/l cADPR.

Supplementary Figure 4. The effect of external NAADP on the  cellstreated with other pyridine nucleotides.  cells preincubated with high glucose concentration were treated with 50 nmol/l (A) and 5 mol/l (B) of NAD, 50 nmol/l (C) and 5 mol/l (D) of NADP, and 50 nmol/l NAADP.

Supplementary Figure 5.NAADP-induced Ca2+ signal was not observed in the presence of 2.8 mmol/l glucose or in the Ca2+-free condition.

Supplementary Figure 6.The 1 mol/l cADPR induced Ca2+ increase was not affected by pretreatment of the cells with bafilomycin A1 or GPN at 37°C for 20 min.

Supplementary Figure 7.The glucose-induced insulin secretionswere not effected by inhibitors of vacuolar H+-ATPase. Islets were preincubated with each inhibitors at 37°C for 20 min or 2 hr in the low glucose condition. Then, after washing to remove already secreted insulin, islets were stimulated with 12 mmol/l glucose containing each inhibitors for 40 min.

Supplementary Figure 8.Ionomycin-induced general Ca2+ increase did not affect cADPR production. Treatment of islets with 12 mmol/l glucose increased the concentration of cADPR, but ionomycin did not show a stimulatory influence on cADPR production. *P < 0.05 versus 2.8 mmol/l glucose.

Supplementary Figure 9. Ten mmol/l KCl-induced Ca2+ increase did not induce any Ca2+ increase by addition of GLP-1 or NAADP.

Supplementary Figure 10. GPN induced Ca2+ release not only from lysosome-containing fractions, but also secretory granule-containing fractions. Ca2+ release from subcellular fractionswas measured to detect Ca2+ movement by GPN using Fluo-3. For subcllular fractions, Ca2+ concentration was measured with 3 mol/l Fluo-3 at 25°C by fluorimeter at 506 nm excitation and 526 nm emission. GPN triggered Ca2+ release not only from lysosome-containing fractions, but also secretory granule-containing fractions. Fraction #1~3 may containsome contaminated lysosomes, because these fractions were -galactosidase-positive (closed triangle) and showed GPN-sensitive Ca2+ release (open circle).

Supplementary Figure 11.The effect of 200 nmol/l insulin on the production of NAADP and cADPR. Islets were preincubated for 40 min with KRB buffer containing 12 mmol/l glucose, and stimulated with insulin for different periods of time. NAADP and cADPR were measured using a cyclic enzymatic assay as described in the manuscript.

SUPPLEMENTARY FIGURE 1

SUPPLEMENTARY FIGURE 2

SUPPLEMENTARY FIGURE 3

SUPPLEMENTARY FIGURE 4

SUPPLEMENTARY FIGURE 5

SUPPLEMENTARY FIGURE 6

SUPPLEMENTARY FIGURE 7

SUPPLEMENTARY FIGURE 8

SUPPLEMENTARY FIGURE 9

SUPPLEMENTARY FIGURE 10

SUPPLEMENTARY FIGURE 11