Isolation of pancreatic islets
Pancreatic islets were isolated from 20-week-old male ICR mice followed the method described previously with modifications [61]. Mice were sacrificed by dislocation and its abdomens were opened with standard scissors. By using a Johns Hopkins Bulldog clamp, the end of the common bile duct before its entrance to small intestine was clamped off. The clamped common bile duct was cannulated by a 30 gauge 0.5 in. needle and injected with 1 mg/ml of collagenase Type V (032-17854, Wako Pure Chemical Industries). After injection, the inflated pancreas was collected and incubated with 2 ml collagenase solution in 37°C shaking water bath for 10 minutes followed by centrifugation (300xg, 4°Cfor 5 minutes). Pancreas pellets were washed with HBSS and undigested tissues were removed with the sieve. Remaining tissues in sizes larger than 70 µm were collected with a cell strainer and stored in RPMI medium (11875, Gibco, USA) supplemented with 10% FBS and 1% P/S. Islets were then collected under microscope for primary culture or fixed with 4% paraformaldehyde (43368, Alfa Aesar) in PBS for wholemount staining.
Cell culture
For primary culture of islet cell, cells were maintained in RPMI 1640 medium (Invitrogen) supplemented with 10% fetal bovine serum 1% P/S at 37°C in a 5% CO2
humidified atmosphere. Cultured medium should be replaced within 2 weeks.
For HEK 293T cell, cells were maintained in DMEM medium (Invitrogen) supplemented with 10% fetal bovine serum and 1 mM sodium pyruvate at 37°C in a 5%
CO2 humidified atmosphere.
Antibodies
Mouse monoclonal Insulin antibody (cell signaling, #8138), Rabbit polyclonal Insulin antibody (cell signaling, #4590), Goat polyclonal CAPS2 antibody (Senta Cruz, sc-65014), SIK2-S901 (provided by Dr. Sheng-Chung Lee, Institute of Molecular Medicine and Clinical Medicine, College of Medicine, National Taiwan University), Mouse monoclonal C-Myc antibody (Senta Cruz, sc-40), Rabbit polyclonal Phospho-Threonine antibody (cell signaling, #9381)
Immunofluorescent staining:
For islet cells grown on glass cover slides, cover slides were picked up carefully
paraformaldehyde for 10 min. The cells were perforated with perforation buffer (0.3%
Triton-X-100 in 1× D-PBS) for 10 min and blocked with blocking buffer (0.2% BSA and 0.1% Triton-X-100 in 1× D-PBS) for over 30 min, then probed with indicated primary antibodies (Insulin (Ms) 1:800, Insulin (Rb) 1:100, CAPS2 1:50, SIK2-S901 1:50, C-Myc 1:50) on 4°Covernight. After that, cover slides were washed with wash buffer (0.1% Triton-X-100 in1× D-PBS)and 1× D-PBS once. Secondary antibodies (Alexa 488-conjugated goat anti-rabbit IgG, Alexa 594-conjugated goat anti-mouse IgG, Alexa 488-conjugated donkey anti-goat IgG, Alexa 594-conjugated donkey anti-rabbit IgG (Invitrogen)) with 1:500 dilution were probed against primary antibodies for 2 hour and cell nucleiwere counterstained with Hoechst 34580 (Invrogen, H21486) for 10 minutes. Cover slides were washed again with1× D-PBS once and mounted with mounting buffer (90% glycerol in 1× D-PBS) for observation.
For mouse pancreas FFPE section, sections were de-paraffinized in xylene for 30mins followed by redydrated in graded ethanol from 100%, 100%, 95%, 70% for each 5 minutes and eventually in desterilized water 5 minutes. Antigen retrieval was performed by boiling sections soaked in 10mM citrate buffer (pH 6.0) with microwave (500w) for 10 minutes. After boiling, sections were cooled down for 20 minutes on ice and blocked with blocking buffer (0.2% BSA and 0.1% Triton-X-100 in 1× D-PBS) for over 30 min. Then the following procedures were same as the above.
For wholemount staining of isolated islets, fixed mouse islets were incubated in primary antibody in staining buffer (2% bovine serum albumin (A9647, Sigma-Aldrich) and 0.2% Triton-X100 (X100, Sigma-Aldrich) in 1× D-PBS) for at least 2 hours in room temperature. After wash with PBS, samples were incubated in secondary antibody diluted staining buffer for 2 hours. Samples were washed again and mounted with PBS.
RNA isolation and RT-PCR
Total RNA was isolated from islets according to the protocol supplied with RNeasy Mini Kit (74104, QIAGEN). Total RNA (1 µg) with 1 µl of oligo (dT), 1 µl of 10 mMdNTP Mix and distilled water in a final reaction volume of 28 µl mixture was heated to 65°C for 5 minutes and incubated on ice for at least 1 minute. This tube was added to the other tube containing 10 µl of 5X RT Buffer, 5 µl of 0.1 M DTT, 1 µl RNaseOUT™ Recombinant RNase Inhibitor (40 units/µl, Cat. no. 10777-019, Invitrogen Life Technologies) and 1 µl of SuperScript™ III RT (200 units/µl, Cat.
no.18080093, Invitrogen Life Technologies) and heated to 50°C for 50 minutes and the reaction was inactivated by heating 85°C for 5 minutes.
Cloning of full length CAPS2
The cDNA sequence encoding the amino acid residues 90 through 1260 of the mouse CAPS2 isoform 7, was amplified by two polymerase chain reaction (PCR) using Phusion High-Fidelity DNA Polymerase (2 U/µL, Cat. no. F-530S, Invitrogen Life Technologies). Plasmid pCMV-3B was used as template. Amplification of CAPS2 (90-565) and CAPS2 (565-1260) uses two pair of oligonucleotide primers：
CAPS2-552-F-BamH1
The underlined sequences primer sequences correspond, respectively, to the BamHI, EcoR1 and Xho1 restriction sites.
PCR was performed in a DNA Engine Peltier Thermal Cycler (BIO-RAD), and the program consisted of a single 98°C denaturation step for 1 min, followed by 35 cycles of denaturation at 98°C for 15 s, annealing at 60°C for 30 s and extension at 72°C for 2 min, followed by a final extension at 72°C for 5 min and ends at 4°C. The PCR
products were run on 1 % agarose gel and then purified using Gel/PCR DNA Fragments Extraction Kit (Cat. no. DF100, Geneaid).
The pCMV-myc-CAPS2wt (90-1260) was constructed in two steps that CAPS2 (565-1260) was cloned into pCMV-3B vector firstly and CAPS2 (90-565) was cloned into pCMV-3B inserted with CAPS2 (565-1260). The final PCR products of CAPS2 (90-565) and CAPS2 (565-1260) were digested with BamH1/EcoR1 and EcoR1/Xho1 enzymes respectively. The pCMV-3B vector was firstly digested with EcoR1/Xho1 and the digested CAPS2 (565-1260) was ligated into the vector using T4 ligase (Invitrogen).
Then the recombinant plasmids were transformed into XL1 cells and the transformed bacteria were incubated on LB plates containing 50 µg/ml kanamycin at 37°C for 18 hr, and colonies containing the insert were identified by colony PCR. A positive clone for each construct was grown overnight in LB medium containing 50 µg/ml kanamycin, and then the plasmids were isolated using a Geneaid™ Mini Plasmid Kit (Cat. no.
PAE40,Geneail) according to the manufacturer’s protocol. After sequencing, pCMV-3B with CAPS2 (565-1260) insertion was used as template and digested by BamH1/EcoR1 enzymes. The digested CAPS2 (90-565) was subcloned into pCMV-myc-CAPS2 (565-1260) by the same procedures.
CAPS2 (1-90) was made from gene synthesis (Omics Bio) and subcloned into pCMV-myc-CAPS2 (90-1260) plasmid. The full-length pCMV-myc-CAPS2(1-1260)
was thus constructed.
Site direct mutagenesis of pCMV-myc-CAPS2(1-1260)
Single amino acid mutations of CAPS2 non-phosphorylatable mutation (T1016A, T1052A, T1231A) were created in pCMV-myc-CAPS2 (90-1260) using quick change site directed mutagenesis method. The mutagenic oligonucleotides (forward) used for T1016A, T1052A, T1231A are
Caps2-T1016A-F (5′-GTCAAAAGAACAAGAGCTGCGTTCGAACTC -3′),
Caps2-T1052A-F (5′-CTAAAAAGCAAAGCGCCAAGCTGTGTGCC -3′),
Caps2-T1231A-F (5′-CACAGACGTCTAGCTGTAGAGGAGGC -3′) respectively. For double amino acid mutations (T1016A+T1052A, T1052A+T1231A, T1016A+T1231A), caps2 T1016A or T1052A or T1231A was used as template to produce another mutagenic site. For triple amino acid mutations (T1016A+T1052A +T1231A), constructs of caps2 with double mutation sites was used as template. PCRs for single, double, triple amino acid mutations using PfuUltra II Fusion HS DNA Polymerase (Cat.
no. 600670, Stratagene) were run for 2 min at 95°C, 18 cycles of 30 s at 95°C and 1 min at 55°C, followed by 4 min 30 s at 68°C. The resulting mutant plasmids were verified by DNA sequencing.
SIK2 constructs
The pCMV-flag-SIK2 wild-type (WT), kinase dead (KD), S587A mutant (non- PKA phosphorylatable mutant), S587D mutant (PKA phospho-mimicking mutant) were kindly provided by Dr. Sheng-Chung Lee (Institute of Molecular Medicine and Clinical Medicine, College of Medicine, National Taiwan University).
Transfection
For HEK 293T cell, transient transfection was performed with LipofectamineTM 2000 (Invitrogen, 11668-027) according to manufacturer’s instructions.
For primary culture of islet cell, transient transfection was performed with LipofectamineTM 2000 (Invitrogen, 11668-027). In a 24-well format, cultured medium is replaced with 450 µl of Opti-MEM. 2.8 µl of LipofectamineTM 2000 are diluted in 25µl of Opti-MEM and incubated for 5 minutes then mixed with 0.5 µg of DNA diluted in 25µl of Opti-MEM. After 45 minutes of incubation, the 100 µl of mixed complexes are added to each well containing cultured islet cells and 450 µl of Opti-MEM. Cells are incubated at 37°C in a CO2 incubator and Opti-MEM should be changed to normal
Immunoprecipitation
Whole cell lysates from transfected HEK-293T cells were extracted by lysing cells with RIPA lysis buffer (20mM Tris/HCl, pH 7.4, 0.15M NaCl, 1mM DTT, 1mM EDTA, 1mM EGTA, 5% glycerol, 0.1% Triton X-100, 1X protease inhibitor (Roche) and 1X phosphatase inhibitor (Roche)) at 4°Cfor 15 min followed by centrifugation at 13000g, 4°Cfor 15 min .Immunoprecipitation was performed by incubating 500ug of protein from cell lysates with anti-Flag M2 magnetic beads (Sigma) at 4°C overnight. The precipitated proteins bound on M2 beads were washed 3 times with TBST (0.1% Tween 20 in 1× TBS) and added with 2X SDS sample buffer. Boiling at 95°C for 5 minutes afterwards dissociates the protein complex before western blot analysis.
In vitro kinase assay
The immunoprecipitated protein complexes were washed with TBST and kinase buffer (20 mMTris, pH 7.8, 10 mM MgCl2, 50 mMNaCl, 1 mM DTT, 1 mM EGTA, protease inhibitor (Roche)). Kinase reaction was initiated by adding ATP 50 µM in 30 µl of kinase buffer then incubated the reaction mixture at 37°C for 30 min. Kinase reaction was terminated by adding 10 µl of 4X sample buffer and boiling at 95°C for 5
min. The level of p-Thr and other indicated proteins were analyzed by Western blot.
Western blot analysis
Proteins was separated by 6% SDS-PAGE and eletrophorestically transferred to polyvinylidenedifluoride (PVDF) membranes (PALL) at 300 mA for 2 hr. Membranes were probed with indicated primary antibody (CAPS2 1:200, Flag 1:200, C-Myc 1:200) at 4°C, overnight. After three times of washing by TBST (0.1% Tween 20 in 1× TBS), horseradish peroxidase-conjugated secondary antibodies were incubated for 2 hour at room temperature. Polypeptide bands were visualized using ECL chemiluminescent substrates (Advansta) and LAS-4000 luminescent image analyzer (Fujifilm).
Insulin ELISA
Insulin concentration was measured by using rat/mouse insulin ELISA kit (Mercodia). 10 µl of diluted samples were loaded into 96-microplate wells coated with antibodies. 100 µl of enzyme conjugate 1X solution was add into each well and incubate on a plate shaker (200 rpm) for 2 hr at room temperature. After incubation, samples in 96-microplate wells were wash with 350 µl of 1X wash buffer for six times. 200 µl of
solution was added. Absorbance at OD 450 nm was detected by ELISA reader.
Results
The expression of SIK2 and CAPS2 in mouse pancreatic islets
In our work, we hypothesize that SIK2 regulates insulin secretion through the phosphorylation of CAPS2 in pancreatic β cells. Before studying the interaction between SIK2 and CAPS2, we want to check their expression in mouse pancreatic islets.
The expression of SIK2 and CAPS2 proteins in islets of Langerhans was studied by immunohistochemistry of mouse pancreatic FFPE slides using SIK2, CAPS2 antibodies.
Insulin antibody was used to mark the location of insulin producing β cells. Our data indicated that SIK2 and CAPS2 were both expressed in mouse pancreatic islets and located mainly in the insulin producing β cells (Figure 1).
Subcellular localization of SIK2 and CAPS2 in mouse pancreatic β cell
Although we can roughly study the expression of SIK2 and CAPS2 in islets from pancreatic FFPE slides, the subcellular localization such as insulin vesicles in β cells were limited to observe. Since the structure of islets of Langerhans is integrated and spheroid, it can be isolated and collected under microscope and each of the islet can directly be fixed and probed by antibodies without tissue processing. The wholemount
observe the original condition of intact islets freshly took out from the model animal (Figure 2). Additionally combined with the use of confocal microscope, subcellular localization of proteins in β cells was easily to be observed.
Thus we performed wholemount immunofluorescent staining on islets isolated from ICR mouse to go deep into the precise localization of SIK2 and CAPS2 in β cells.
CAPS2 are cytosolic proteins that also peripherally associated with the membrane of large dense core vesicles (LDCV) in chromaffin cells [62]. Our data showed CAPS2 was expressed in the cytosol and formed vesicle-liked dots which were co-localized with insulin vesicles (Figure 3B). This result suggested in the case of β cells, CAPS2 was associated with insulin-containing LDCVs. However, SIK2 did not show vesicle-liked structures but diffusely expressed in the cytosol of β cells. Similarly, there were no enriched SIK2 signals found at CAPS2 vesicles (Figure 3A).
Cloning of β cells expressed CAPS2 isoform
There are 7 alternative splicing variants in mouse CAPS2 (CAPS2 v1~v7) (Reference to NCBI). CAPS2 v1, which lacks exon 25, represents the longest transcript and encodes the longest isoform. CAPS2 v2 is an isoform lacking exons 22 and 25, which encode part of the Munc13-1-homologous domain (MHD). CAPS2 v3 lacks
exons 11 and 22. CAPS2 v4, 5, 6 have C-terminal deletions from exon 14, exon 12, and exon 5, respectively. CAPS2 v7 lacks exons 17, 19 and 22 (Figure 4A). These isoforms can be subdivided into two groups: a long form containing the C-terminal MHD and a short form lacking the C-terminal MHD. CAPS2 v1, 2, 3 and 7 belong to the long-form group, while CAPS2 v4, 5 and 6 belong to the short-form group [63].
Different alternative splicing variants have distinct expression and functional properties. In order to investigate the role of CAPS2 and its interaction with SIK2 in pancreatic β cells, firstly we wanted to identify the β cells expressed CAPS2 splicing variants. As the MHD domain is involved in calcium dependent exocytosis and the C-terminal region is required for DCV (dense-core vesicle) binding, we target long-form group of CAPS2, which includes MHD domain and C-terminal region and may possess more complete functions. As a result, we designed primers by sequences of 5' end and 3' C-terminal region trying to clone it from isolated mouse pancreatic islets.
The cDNA sequence encoding the mouse CAPS2 variant 7 was amplified and cloned into pCMV-3B vector (Figure 4B). The pCMV-CAPS2 v7-WT (wild-type) construct was transfected into HEK 293T cells to confirm if cells could correctly express the myc-tagged CAPS2 recombinant protein. Data from western blots showed HEK 293T cells could overexpress the exogenous myc-tagged CAPS2 recombinant protein probed by CAPS2 and myc antibodies and its molecular weight was larger than the endogenous
CAPS2 protein (Figure 4C). Also, immunofluorescent staining of HEK 293T cells transfected with this construct again confirmed the correct expression of exogenous myc-tagged CAPS2 recombinant protein in the cytosol (Figure 4D).
Physiological relevant proteins interaction between SIK2 and CAPS2
To study the correlation between SIK2 and CAPS2, we check their physiological relevant proteins interaction for a start by co-immunoprecipitation technique, a way using target protein-specific antibodies to indirectly capture proteins that are bound to a specific target protein. Furthermore, we wanted to experiment if their physiological relevant proteins interaction was in relation with SIK2's kinase activity. It has been reported that SIK2 suppressed CREB (cAMP-response element binging protein) mediated gene expression by phosphorylating its co-activator, TORC2. Nevertheless, cAMP-PKA dependent phosphorylation of SIK2 at Ser-587 diminished its TORC2 phosphorylation [42]. We therefore used SIK2-S587A mutant, the non-PKA phosphorylatable mutant, to mimic kinase active form and SIK2-S587D mutant, the PKA phosphorylating-mimicking mutant, to mimic kinase inactive form.
HEK 293T cells were co-transfected with pCMV-myc-CAPS2-WT construct as well as pCMV-flag-SIK2 constructs including Mock, SIK2-WT (wild-type),
SIK2-S587A (mimicked kinase active form), SIK2-S587D (mimicked kinase inactive form), SIK2-KD (kinase dead) respectively. Flag-SIK2 recombinant proteins were immunoprecipitated from the transfected HEK 293T cell lysates by a flag antibody and analyzed whether CAPS2 is its binging partner by western blot.
According to our data, CAPS2 was presented in the captured SIK2 protein complex and the amount of bound CAPS2 is associated with SIK2 kinase activity.
CAPS2 proteins tended to bind SIK2 without kinase activity (SIK2-KD) active rather than wild-type SIK2 (SIK2-WT). In addition, the phosphorylation on SIK2 by PKA also affected its protein binding with CAPS2. However, both the non-PKA phosphorylatable mutantSIK2-S587A, thought as a SIK2 kinase active form,and the PKA phosphorylating-mimicking mutantSIK2-S587D, which mimics SIK2 kinase inactive form, showed increased levels on CAPS2 binding compared to wild-type SIK2 (SIK2-WT) (Figure 5) .
Putative SIK2 phosphorylation sites on CAPS2 and CAPS2-MUT-T1016A, T1052A, T1231A construct
After checking the physiological protein-protein interaction between SIK2 and CAPS2, we attempted to verify if CAPS2, the putative substrate of SIK2, would be
directly phosphorylated by SIK2. SIK2 is a serine/threonine kinase and its putative phosphorylation sites on CAPS2 predicted using by Scansite in silico database mining (http://scansite.mit.edu) were mostly on threonine residue. These putative target sites were scored and the lower the numbers were, the more optimal the target sites were.
Threonine residues located at 1016, 1052, 1231 of CAPS2, which scored 0.00 could be the most potential SIK2 phosphorylation sites (Figure 6A). To further experiment whether SIK2 phosphorylates CAPS2 right on these 3 position, we made pCMV-myc-CAPS2-MUT (T1016A, T1052A, T1231A) construct by performing site direct mutagenesis via the replacement of threonine with alanine, which is the non-phosphorylatable mutation on pCMV-myc-CAPS2-WT construct at T1016, T1052, T1231 (Figure 6B).
In vitro phosphorylation of SIK2 on CAPS2
We study enzymatic phosphorylation ability of SIK2 on CAPS2 by performing in vitro kinase assay, a method which puts a kinase and suspicious substrate together in a
kinase reacting condition to see if the phosphorylation happens in vitro. HEK 293T cells were transfected with pCMV-myc-CAPS2-WT/MUT and pCMV-flag-SIK2- Mock/WT/S587A/S587D/KD constructs individually, and the expressed recombinant
proteins were captured form cell lysates by immunoprecipitation. CAPS2-WT and CAPS2-MUT proteins respectively were put together with SIK2 proteins owning different kinase activities for proceeding kinase reaction followed by western blot analysis. As the predicted SIK2 putative phosphorylation sites on CAPS2 were mostly on threonine residue, we used specific p-Thr antibody to detect p-Thr (phospho-threonine) level of CAPS2. Data revealed that SIK2 indeed phosphorylated CAPS2 at threonine residue. Since there was no signals from SIK2-KD, indicating the p-Thr signal of CAPS2 was truly came from SIK2 kinase activity. The p-Thr level of CPAS2-MUT (T1016A, T1052A, T1231A) was slightly decreased compared to CPAS2-WT, indicating that SIK2 would phosphorylate CAPS2 not only at T1016, T1052, T1231 but threonine at other position. Surprisingly, both the non-PKA phosphorylatable mutant SIK2-S587A and the PKA phosphorylating-mimicking mutant SIK2-S587D showed reduced kinase activity, suggesting SIK2-S587 is crucial for the regulation of SIK2 kinase activity. However there were not much differences in the CAPS2 phosphorylation level by SIK2-S587A or SIK2-S587D. (Figure 7).
Transfection and expression of pTimer-phogrin in primarily cultured islet cells
After verifying the interaction between SIK2 and CAPS2, we desired to reveal how
their cross-talk took part in insulin secretion in pancreatic β cells. To do so, we tried to conduct functional analysis of these two molecules in our model system, the primarily cultured β cells of isolated islets from ICR mice, which was close the most to physiological states instead of cell lines. These primary β cells attached to the bottom of culture dish and spread out after more than 3 weeks of culture. The capabilities of primary β cells to produce and secrete insulin were tested by immunofluorescent staining, suggesting primary β cells retained these abilities after long-term culture (Figure 8A).
We next tried to perform gene overexpression in primary β cells with lipofectamine transfection. Phogrin is a transmembrane protein expressed in pancreatic β cells, in which it is localized to the membrane of insulin-containing dense-core vesicles [64]. We delivered the pTimer-phogrin construct into primary β cells followed by immunocytochemistry targeting insulin. Results showed signals of pTimer-phogrin appeared to be granules which co-localized with most insulin vesicles, revealing transfection and expression with our procedures worked well in primarily cultured islet cells (Figure 8B).
Insulin secretion of SIK2/CAPS2 overexpressed mouse primary β cells
Followed by successful establishment of gene overexpression in primary β cells, we could begin to overexpress our purposed molecules and study their functions in β cells. Isolated pancreatic islets were collected and seeded separately under microscope into 24-well culture plate asa single islet seeded into each well. After a 3-week-culture, the attached islets were transfected with Mock, CAPS2, SIK2, and SIK2 plus CAPS2 constructs following the modified protocol for lipofectamine transfection. The next day after transfection, islets were starved for 30 min by the replacement with warmed HBSS before 16.7 mM of glucose stimulation. Solutions were collected at the beginning and at the end of starvation and every 15 min after glucose stimulation. The amounts of insulin secreted into solutions were measured by mouse insulin ELISA. In groups transfected with Mock, CAPS2, SIK2, and SIK2 plus CAPS2, we had 4, 2, 3 and 3 islets,
Followed by successful establishment of gene overexpression in primary β cells, we could begin to overexpress our purposed molecules and study their functions in β cells. Isolated pancreatic islets were collected and seeded separately under microscope into 24-well culture plate asa single islet seeded into each well. After a 3-week-culture, the attached islets were transfected with Mock, CAPS2, SIK2, and SIK2 plus CAPS2 constructs following the modified protocol for lipofectamine transfection. The next day after transfection, islets were starved for 30 min by the replacement with warmed HBSS before 16.7 mM of glucose stimulation. Solutions were collected at the beginning and at the end of starvation and every 15 min after glucose stimulation. The amounts of insulin secreted into solutions were measured by mouse insulin ELISA. In groups transfected with Mock, CAPS2, SIK2, and SIK2 plus CAPS2, we had 4, 2, 3 and 3 islets,