Regulation of PKD by the MAPK p38δ in Insulin Secretion and Glucose Homeostasis

Summary Dysfunction and loss of insulin-producing pancreatic β cells represent hallmarks of diabetes mellitus. Here, we show that mice lacking the mitogen-activated protein kinase (MAPK) p38δ display improved glucose tolerance due to enhanced insulin secretion from pancreatic β cells. Deletion of p38δ results in pronounced activation of protein kinase D (PKD), the latter of which we have identified as a pivotal regulator of stimulated insulin exocytosis. p38δ catalyzes an inhibitory phosphorylation of PKD1, thereby attenuating stimulated insulin secretion. In addition, p38δ null mice are protected against high-fat-feeding-induced insulin resistance and oxidative stress-mediated β cell failure. Inhibition of PKD1 reverses enhanced insulin secretion from p38δ-deficient islets and glucose tolerance in p38δ null mice as well as their susceptibility to oxidative stress. In conclusion, the p38δ-PKD pathway integrates regulation of the insulin secretory capacity and survival of pancreatic β cells, pointing to a pivotal role for this pathway in the development of overt diabetes mellitus.

five times to the C57BL/6 background and intercrossed to generate homozygous knockout mice.
Cell culture, transfection and cell sorting. Transfection of 293T cells was performed using calcium phosphate precipitation. INS1 cells were transfected with the pcDNA3.1-HA-p38δ, pcDNA3.1-HA-p38δ F324S expression plasmid or the empty vector using Fugene reagent (Roche) and selected with 500 μg/ml neomycin (Sigma). Knockdown of p38δ in MIN6 was obtained by lentivirus-mediated transduction of short hairpin against p38δ RNA (sequences are provided in supplemental experimental procedures) and selection with puromycin (4 μg/ml). Transient knockdown of PKD1 in INS1 or MIN6 cells was performed by electroporation. Briefly, cells were resuspended in OPTImem (Invitrogen) and electroporated at 270 V and 960 µF with 2 μM of siRNA duplex using the Gene Pulser Xcell (Bio-Rad). Forty-eight hours after electroporation, cells were subjected to analysis as described below. Transfection of INS1 cells with GFP or GFP-tagged PKD WT or mutant forms was achieved by electroporation as described above. To enrich for GFP-positive cells, INS1 cells were sorted 24 hours after transfection using FACS Aria Cell Sorter (BD) equipped with an argon laser emission of 488 nm. Non-transfected cells were used as a control to set the cutoff value for background fluorescence.

LC-MS/MS analysis.
The samples were analyzed on a hybrid LTQ-FTICR mass spectrometer (Thermo, San Jose, CA) interfaced with a nanoelectrospray ion source. Chromatographic separation of peptides was achieved on an Eksigent nano LC system (Eksigent Technologies, Dublin, CA, USA), equipped with a 11 cm fused silica emitter, 75 μm inner diameter (BGB Analytik, Böckten, Switzerland), packed inhouse with a Magic C18 AQ 5 μm resin (Michrom BioResources, Auburn, CA, USA). Peptides were loaded from a cooled (4°C) Spark Holland auto sampler and separated using ACN/water solvent system containing 0.1 % formic acid with a flow rate of 300 nl/min. Peptide mixtures were separated with a gradient from 3 to 30 % ACN in 90 min. Up to three data-dependent MS2 spectra were acquired in the linear ion trap for each FT-MS spectral acquisition range, the latter acquired at 100,000 FWHM nominal resolution settings with an overall cycle time of approximately 1 s. Charge state screening was employed to select for ions with at least two charges and rejecting ions with undetermined charge state. For each peptide sample, a standard data-dependent acquisition method on the three most intense ions per MS-scan was used and a threshold of 200 ion counts was used for triggering an MS2 attempt.

LC-MS/MS data analysis.
The MS2 data were, dependant on the sample analyzed, searched against the human (v3.23) and mouse IPI database (v3.26) non redundant database using SORCERER-SEQUEST(TM) (Eng et al., 1994) v3.0.3, which was run on the SageN Sorcerer2 (Thermo Electron, San Jose, CA, USA). For the in silico digest, trypsin was defined as protease, cleaving after K and R (if followed by P the cleavage was not allowed). Two missed cleavages and one non-tryptic terminus were allowed for the peptides that had a maximum mass of 6000 Da. The precursor ion tolerance was set to 25 p.p.m. The data were searched allowing phosphorylation (+79.9663 Da) of serine, threonine and tyrosine as a variable modification and carboxyamidomethylation of cysteine (+57.0214 Da) residues as a fixed modification. In the end, the search results obtained bySequest were subjected to statistical filtering using PeptideProphet (Keller et al., 2002) (v3.0) and ProteinProphet (v3.0) (Keller et al., 2005). The phosphopeptide tandem mass spectra were furthermore manually inspected and tested for the presence of a neutral loss peak of -98 Da which is indicative for a phosphopeptides under CID (Bodenmiller et al., 2007). Finally, also the MS1 mass shift between the non-phosphopeptide and phosphopeptide of +79.9663 Da was used to verify the results.

Transmission electron microscopy (TEM).
To evaluate the type, abundance and distribution of secretory granules, sections of p38δ +/+ and p38δ Δ/Δ mice were prepared for conventional electron microscopy. Quantitative analyses of pancreatic β cells have been assessed by TEM as previously described (Stefan et al., 1987). Briefly, the areas of cytoplasm, Golgi apparatus and peripheral membrane compartment (defined as a band of 300 nm thickness, about the diameter of one secretory granule, along the cell membrane) were evaluated by semi-automatic planimetry of photographs taken at a magnification of 19000x, using a graphic tablet and the Leica Qwin software (Leica, Glattbrugg, Switzerland). The areas of secretory granules in these 3 compartments and the length of cell membrane were similarly evaluated. The number of mature (small very electron dense core, large electron lucent halo) and immature granules (medium dense core, thin peripheral halo) was further scored in all compartments. From these data, the volume density of secretory ganules was calculated in each β-cell by dividing the cumulated areas of these organelles by that of the cytoplasm, Golgi or membrane compartment, respectively. The numerical density of granules apposed to the cell membrane was given by dividing the number of granules scored in the membrane compartment by the length of the cell membrane. The proportion of mature and immature granules in each compartment was given by dividing the number of each granule type by the total number of granules. Data were expressed as mean ± SEM and compared by either analysis of variance and ad Scheffe's ad hoc t-test (volume and numerical density of granules) or by the Chi-square test (distribution of secretory granules), as provided by the SPSS software (SPSS, Chicago, USA).

Primers and shRNA sequences. Name
Sequence

Supplemental References
Bodenmiller    Quantitative histological assessment of total islet area in relation to total pancreatic area. No significant differences in p38δΔ/Δ (black bars, n=3) compared to p38δ+/+ (white bars, n=3) could be observed. (C) Total insulin extracted from pancreas measured by radio-immuno assay. No significant difference in total insulin content (μg/mg) was detected between p38δΔ/Δ (black bars, n=3) and p38δ+/+ mice (white bars, n=3). (D) The mean insulin content per isolated islet was also similar in p38δΔ/Δ (black bars, n=14) and p38δ+/+ mice (white bars, n=16). All error bars indicate ±SEM. i, peak glucose response, plateau glucose response, and the peak response to KCl. No significant difference could be observed between p38δ+/+ and p38δΔ/Δ islets. (C) Representative Ca2+ current in isolated pancreatic islets of p38δ+/+ (black solid line, n=20) and p38δΔ/Δ (grey solid line, n=17) mice evoked by a depolarisation from -70 mV to 0 mV. Currents are shown together with respective means ±SEM of the peak and sustained currents (measured at the end of the pulse). No significant differences was observed between p38δ+/+ and p38δΔ/Δ islets. (D) Summary of the average charge entry (derived from the area under the curve during the depolarisation) during the 10ms pulse from -70 mV to 0 mV in p38δ+/+ (+/+) and p38δΔ/Δ (Δ/Δ) β cells. No significant difference was observed between p38δ+/+ and p38δΔ/Δ islets. All error bars indicate ±SEM.    Western blotting with total cell lysates from pancreas of p38δ+/+ and p38δΔ/Δ mice revealed an enhanced autophosphorylation of PKD (Serine 916). Tubulin was used to confirm equal loading. Figure S9. p38δ deletion leads to dispersion of furin convertase and GM130 in primary pancreatic β cells. p38δ-deficient (p38δΔ/Δ) and wild type (p38δ+/+) pancreatic β cells were analyzed by immunofluorescence microscopy using antibody against Golgi-proteins furin convertase (furin, red) and GM130 (green). DNA was counterstained with DAPI (blue). Boxes in the left column outline the areas that have been magnified in the panels below. Size bar 10 μm. Figure S10. p38δ deletion leads to dispersion of giantin in MIN6 cells. MIN6 cells stably expressing shRNA against p38δ (p38δ shRNA) and cells expressing the empty vector (vector control) were cultured and analyzed by immunofluorescence microscopy using antibodies against giantin (red). DNA was stained with DAPI (blue). Note that p38δ shRNA expressing cells show dislocalization of giantin (arrows). Size bar 10 μm.

Figure S12. Inactivation of PKD1 and p38δ in MIN6 cells and stable expression of PKD1 mutants in INS1 cells. (A)
Western blot with total cell lysates from MIN6 cells stably expressing shRNA against p38δ (shRNA p38δ) or a control vector (vector control) transfected with scrambled control siRNA (siRNA scramb.) or siRNA against PKD1 (siRNA PKD1). Knockdown by siRNA was about 60%. Tubulin WB indicated equal loading (lower panel). (B) Western blot with total cell lysates from INS1 cells overexpressing GFP or GFP-tagged WT and mutant forms of PKD1. GFP positive cells were enriched by FACS sorting with a purity of 95% before conducting experiments (data not shown). Tubulin WB indicated equal loading (lower panel).

Figure S13. Reduced weight gain and attenuated insulin resistance in p38δΔ/Δ compared to p38δ+/+ mice on a high-fat diet. (A)
Insulin tolerance tests. Insulin (1 U/kg) was injected intraperitoneally in ad libitum fed mice and glucose was measured at indicated time points. Insulin sensitivity was equal in p38δΔ/Δ (squares and dotted line, n=7) and p38δ+/+ (diamonds and solid line, n=5) mice on a normal diet. However, insulin sensitivity was significantly improved in p38δΔ/Δ (n=11) compared to p38δ+/+ mice (n=6) on a HF diet (*p<0.05 and **p<0.01). (B) Body weights of p38δΔ/Δ and p38δ+/+ mice on a high-fat (HF, n=11 and n=6, respectively) or normal diet (ND, n=7 and n=5, respectively) were determined at indicated time points. p38δΔ/Δ mice gained less weight on a HF diet compared to p38δ+/+ (*p<0.05), while no differences could be observed on a normal diet. All error bars indicate ±SEM. Figure S14. Islet growth under high-fat feeding conditions is similar in p38δΔ/Δ and p38δ+/+ mice. Quantitative histological assessment of total islet area in relation to total pancreatic area. Islet growth in response to high-fat feeding (HF) in p38δ+/+ (+/+, n=3) mice was significantly enhanced compared to mice on a normal diet (ND). The islet area of p38δΔ/Δ (Δ/Δ, n=3) mice on a high-fat diet was also increasing compared to mice on a normal diet but differences were not significant. However, no signifcant (ns) difference in the islet area between p38δ+/+ and p38δΔ/Δ mice on a high-fat diet could be observed. All error bars indicate ±SEM.  Values are mean ± SEM of the indicated number of β-cells.