• 沒有找到結果。

In this study, we provided evidences that CAPS2 was the downstream substrate of SIK2. On one hand, from the physiological protein interaction between SIK2 and CAPS2, we found CAPS2 was prone to associate with SIK2 without kinase activity (SIK2-KD) instead of the kinase-active SIK2 (SIK2-WT), suggesting the affinity between SIK2 and CAPS2 was lowered after the phosphorylation of CAPS2 by SIK2.

This result may be due to the conformational change of CAPS2 after getting phosphorylated. In other words, SIK2 tends to bind with unphosphorylated CAPS2.

On the other hand, our results from in vitro kinase assay demonstrated SIK2 indeed phosphorylates CAPS2 at threonine residue. Additionally, it has been reported that cAMP-PKA dependent phosphorylation of SIK2 at Ser-587 diminished phosphorylation of TORC2 by SIK2, which suppressed CREB (cAMP-response element binging protein) mediated gene expression. Our previous data indicated that SIK2 was phosphorylated by PKA on Ser587 upon glucose stimulation, which promoted insulin secretion.

Equivalently, inhibition of SIK2 kinase activity by treating with selectively SIK2 kinase inhibitor (0.3 µM Compound C) increased insulin secretion. We thus assumed SIK2-S587 as kinase inactive form because it promoted insulin secretion. However, we discovered both non-PKA phosphorylatable SIK2-S587A mutant, which mimics SIK2

SIK2 kinase inactive form, compromised SIK2 kinase activity compared to SIK2-WT.

These results suggested SIK2-S587 is crucial for the regulation of SIK2 kinase activity as different substitutions at this position may result in the conformational changes and thus reduced its kinase function, which explained why SIK2-S587A and SIK2-S587D showed increased levels of SIK2-CAPS2 binging. That is to say, when kinase activity of SIK2 is restrained, it favors the formation of the SIK2-CAPS2 complex.

We performed gene overexpression in primarily cultured β cells to study how SIK2/CAPS2 crosstalk participates in insulin secretion. β cells co-expressed with both SIK2 and CAPS2, in which more SIK2-CAPS2 complexes may form, have strengthened responses to glucose stimulation (0~15 min). SIK2-CAPS2 complexes sensitize the rapid glucose stimulated insulin release possibly because it triggers the release of docked insulin vesicles.

Altogether, based on our results, we proposed a model (Figure 10) to elaborate how SIK2 regulates insulin secretion through binding with and phosphorylating on CAPS2. In basal state, SIK2 phosphorylates CAPS2 thereby interfere the binding between SIK2 and CAPS2. Upon glucose stimulation, the increased SIK2-S587 which was phosphorylated by cAMP dependent PKA pathway diminished its kinase activity therefore more SIK2-CAPS2 complexes could be composed. Consequently, releasing of docked insulin vesicles are promoted. Other possibilities are emerged that there are

probably other molecules take part in modulating phosphorylation and de-phosphorylation dynamics of CAPS2.

Nevertheless, the three most potential SIK2 phosphorylation site predicted on CAPS2 at T1016, T1052 and T1231 are not necessary for SIK2 phosphorylation because replacement of T with non-phosphorylatable A showed little decrease in p-Thr level. There were also other predicted phosphorylation site with higher probabilities on the mouse β cells expressed CAPS2 such as T434, T571, T793, T828.

Since different splicing variants of a protein may present different expression pattern and distribution, we therefore cloned β cells expressed CAPS2 splicing variants from mouse isolated islets. Sadakata et al in 2007 reported the differential expression and functional properties of six splicing variants in mouse CAPS2 that CAPS2 v1 was expressed exclusively in the brain, and CAPS2 v2~6 were highly expressed in the brain, but also in some non-neural tissues. Moreover in the brain, all isoforms showed predominant expression patterns in the cerebellum. CAPS2 v1 showed an up-regulated expression pattern in the developing cerebellum, whereas CAPS2 v2~6 exhibited transiently peaked expression patterns. CAPS2 v3 which lack the PH domain is important for membrane association, showed slightly decreased BDNF-releasing activity. In our experiment, CAPS2 variant 7 (Reference to NCBI), which lacks exons 17, 19 and 22, and encode part of the Munc13-1-homologous domain (MHD), was

identified in mouse pancreatic β cells. However, the expression, distribution and function of mouse CAPS2 variant 7 has yet been documented. The particular function of CAPS2 variant 7 and its interaction with SIK2 in pancreatic β cells are required more investigation.

CAPS1 and CAPS2 have been shown to regulate Ca2+-dependent exocytosis of dense-core vesicle release of transmitters and hormones in neuroendocrine cells, but their precise roles in the secretory process still remain elusive. Speidel et al in 2007 pointed out the essential role of CAPS1 and CAPS2 in pancreatic β cells by generating CAPS2-/-, and CAPS1+/- ; CAPS2-/-mice, presenting glucose intolerance which is attributable to a marked reduction of glucose-induced insulin secretion. This phenomenon correlates with diminished Ca2+-dependent exocytosis, a reduction in the size of the morphologically docked pool, and a decrease in the readily releasable pool of secretory vesicles. Accordingly, our proposed SIK2/CAPS2 model may support this study that knockout of CAPS2 cause inability to form SIK2-CAPS2 complexes contributed to the decreasing glucose stimulated insulin secretion.

For treatment of type 2 diabetes clinically, metformin and thiazolidinediones (TZDs), examples of the widely used drug decrease glucose production from liver and increase glucose utilization by skeletal myocyte, through the indirect activation of AMPK. AMPK is an energy-sensing enzyme that is activated when cellular energy

levels are low, and it signals to stimulate glucose uptake in skeletal muscles, fatty acid oxidation in adipose (and other) tissues, and reduces hepatic glucose production. AMPK activation improves insulin sensitivity and glucose homeostasis, making it an attractive target for T2D and metabolic syndrome. On the other side, insulin signaling pathways in peripheral tissues become defective under the circumstances of insulin resistance; as a result, the need for insulin is further enhanced. Increasing insulin secretion in order to compensate for insulin resistance is also essential way for treatment of type 2 diabetes.

Our results here show that the cross talk between SIK2 and CAPS2 in β cells regulates insulin secretion responsed to glucose stimulation. Further study to reveal the molecular machineries of their interaction can potentially be applicable to develop novel medical strategies to confer symptoms of patients with type 2 diabetes.

References

1.   Mehlen,  P.  and  A.  Puisieux,  Metastasis:  a  question  of  life  or  death.  Nat  Rev   Cancer,  2006.  6(6):  p.  449-­‐58.  

2.   Stewart,  B.W.,  et  al.,  World  cancer  report  2014.  xiv,  630  pages.  

3.   Jemal,  A.,  et  al.,  Cancer  statistics,  2004.  CA  Cancer  J  Clin,  2004.  54(1):  p.  8-­‐29.  

4.   Mazzone,  P.,  et  al.,  Bronchoscopy  and  needle  biopsy  techniques  for  diagnosis   and  staging  of  lung  cancer.  Clin  Chest  Med,  2002.  23(1):  p.  137-­‐58,  ix.  

5.   Velve-­‐Casquillas,  G.,  et  al.,  Microfluidic  tools  for  cell  biological  research.  Nano   Today,  2010.  5(1):  p.  28-­‐47.  

6.   Velve  Casquillas,  G.,  et  al.,  Fast  microfluidic  temperature  control  for  high   resolution  live  cell  imaging.  Lab  Chip,  2011.  11(3):  p.  484-­‐9.  

7.   Velve-­‐Casquillas,  G.,  et  al.,  A  fast  microfluidic  temperature  control  device  for   studying  microtubule  dynamics  in  fission  yeast.  Methods  Cell  Biol,  2010.  97:  p.  

185-­‐201.  

8.   Forry,  S.P.  and  L.E.  Locascio,  On-­‐chip  CO2  control  for  microfluidic  cell  culture.  

Lab  Chip,  2011.  11(23):  p.  4041-­‐6.  

9.   Lu,  P.,  V.M.  Weaver,  and  Z.  Werb,  The  extracellular  matrix:  a  dynamic  niche  in   cancer  progression.  J  Cell  Biol,  2012.  196(4):  p.  395-­‐406.  

10.   Cox,  T.R.  and  J.T.  Erler,  Remodeling  and  homeostasis  of  the  extracellular  matrix:  

implications  for  fibrotic  diseases  and  cancer.  Dis  Model  Mech,  2011.  4(2):  p.  

165-­‐78.  

11.   Jia,  D.,  et  al.,  Development  of  a  highly  metastatic  model  that  reveals  a  crucial   role  of  fibronectin  in  lung  cancer  cell  migration  and  invasion.  BMC  Cancer,  2010.  

10:  p.  364.  

12.   Sun,  X.,  et  al.,  The  EDA-­‐containing  cellular  fibronectin  induces  

epithelial-­‐mesenchymal  transition  in  lung  cancer  cells  through  integrin   alpha9beta1-­‐mediated  activation  of  PI3-­‐K/AKT  and  Erk1/2.  Carcinogenesis,   2014.  35(1):  p.  184-­‐91.  

13.   Quail,  D.F.  and  J.A.  Joyce,  Microenvironmental  regulation  of  tumor  progression   and  metastasis.  Nat  Med,  2013.  19(11):  p.  1423-­‐37.  

14.   Yeh,  H.Y.  and  J.C.  Lin,  Surface  characterization  and  in  vitro  platelet  compatibility   study  of  surface  sulfonated  chitosan  membrane  with  amino  group  

protection-­‐deprotection  strategy.  J  Biomater  Sci  Polym  Ed,  2008.  19(3):  p.  

291-­‐310.  

15.   Chen,  Y.H.,  et  al.,  Control  of  cell  attachment  on  pH-­‐responsive  chitosan  surface   by  precise  adjustment  of  medium  pH.  Biomaterials,  2012.  33(5):  p.  1336-­‐42.  

16.   Chen,  Y.H.,  et  al.,  Cell  fractionation  on  pH-­‐responsive  chitosan  surface.  

Biomaterials,  2013.  34(4):  p.  854-­‐63.  

17.   Chu,  Y.W.,  et  al.,  Selection  of  invasive  and  metastatic  subpopulations  from  a   human  lung  adenocarcinoma  cell  line.  Am  J  Respir  Cell  Mol  Biol,  1997.  17(3):  p.  

353-­‐60.  

18.   American  Diabetes,  A.,  Diagnosis  and  classification  of  diabetes  mellitus.  

Diabetes  Care,  2011.  34  Suppl  1:  p.  S62-­‐9.  

19.   Classification  and  diagnosis  of  diabetes  mellitus  and  other  categories  of  glucose   intolerance.  National  Diabetes  Data  Group.  Diabetes,  1979.  28(12):  p.  1039-­‐57.  

20.   Daneman,  D.,  Type  1  diabetes.  Lancet,  2006.  367(9513):  p.  847-­‐58.  

21.   Butler,  A.E.,  et  al.,  Beta-­‐cell  deficit  and  increased  beta-­‐cell  apoptosis  in  humans   with  type  2  diabetes.  Diabetes,  2003.  52(1):  p.  102-­‐10.  

22.   American  Diabetes,  A.,  Standards  of  medical  care  in  diabetes-­‐-­‐2014.  Diabetes   Care,  2014.  37  Suppl  1:  p.  S14-­‐80.  

23.   Krupp,  M.A.,  et  al.,  Current  medical  diagnosis  &  treatment.  McGraw-­‐Hill   Companies:  New  York  etc.  ,.  p.  v.  

24.   Diabetes  care.  Med  Econ,  2004.  81(17):  p.  14,  16.  

25.   Jiang,  Y.D.,  et  al.,  Incidence  and  prevalence  rates  of  diabetes  mellitus  in  Taiwan:  

analysis  of  the  2000-­‐2009  Nationwide  Health  Insurance  database.  J  Formos   Med  Assoc,  2012.  111(11):  p.  599-­‐604.  

26.   Seino,  S.  and  G.I.  Bell,  Pancreatic  beta  cell  in  health  and  disease.  2008,  Tokyo:  

Springer.  xv,  474  p.  

27.   Ma,  L.,  et  al.,  Direct  imaging  shows  that  insulin  granule  exocytosis  occurs  by   complete  vesicle  fusion.  Proc  Natl  Acad  Sci  U  S  A,  2004.  101(25):  p.  9266-­‐71.  

28.   Ammala,  C.,  et  al.,  Exocytosis  elicited  by  action  potentials  and  voltage-­‐clamp   calcium  currents  in  individual  mouse  pancreatic  B-­‐cells.  J  Physiol,  1993.  472:  p.  

665-­‐88.  

29.   Rorsman,  P.  and  E.  Renstrom,  Insulin  granule  dynamics  in  pancreatic  beta  cells.  

Diabetologia,  2003.  46(8):  p.  1029-­‐45.  

30.   Neher,  E.,  Vesicle  pools  and  Ca2+  microdomains:  new  tools  for  understanding   their  roles  in  neurotransmitter  release.  Neuron,  1998.  20(3):  p.  389-­‐99.  

31.   Parsons,  T.D.,  et  al.,  Docked  granules,  the  exocytic  burst,  and  the  need  for  ATP   hydrolysis  in  endocrine  cells.  Neuron,  1995.  15(5):  p.  1085-­‐96.  

32.   Gromada,  J.,  et  al.,  CaM  kinase  II-­‐dependent  mobilization  of  secretory  granules   underlies  acetylcholine-­‐induced  stimulation  of  exocytosis  in  mouse  pancreatic   B-­‐cells.  J  Physiol,  1999.  518  (  Pt  3):  p.  745-­‐59.  

33.   Barg,  S.,  et  al.,  Fast  exocytosis  with  few  Ca(2+)  channels  in  insulin-­‐secreting   mouse  pancreatic  B  cells.  Biophys  J,  2001.  81(6):  p.  3308-­‐23.  

34.   Rorsman,  P.,  et  al.,  The  Cell  Physiology  of  Biphasic  Insulin  Secretion.  News  

Physiol  Sci,  2000.  15:  p.  72-­‐77.  

35.   Olofsson,  C.S.,  et  al.,  Fast  insulin  secretion  reflects  exocytosis  of  docked  granules   in  mouse  pancreatic  B-­‐cells.  Pflugers  Arch,  2002.  444(1-­‐2):  p.  43-­‐51.  

36.   Cheng,  K.,  S.  Andrikopoulos,  and  J.E.  Gunton,  First  phase  insulin  secretion  and   type  2  diabetes.  Curr  Mol  Med,  2013.  13(1):  p.  126-­‐39.  

37.   Seino,  S.,  T.  Shibasaki,  and  K.  Minami,  Dynamics  of  insulin  secretion  and  the   clinical  implications  for  obesity  and  diabetes.  J  Clin  Invest,  2011.  121(6):  p.  

2118-­‐25.  

38.   Wang,  Z.,  et  al.,  Cloning  of  a  novel  kinase  (SIK)  of  the  SNF1/AMPK  family  from   high  salt  diet-­‐treated  rat  adrenal.  FEBS  Lett,  1999.  453(1-­‐2):  p.  135-­‐9.  

39.   Horike,  N.,  et  al.,  Adipose-­‐specific  expression,  phosphorylation  of  Ser794  in   insulin  receptor  substrate-­‐1,  and  activation  in  diabetic  animals  of  salt-­‐inducible   kinase-­‐2.  J  Biol  Chem,  2003.  278(20):  p.  18440-­‐7.  

40.   Lin,  X.,  et  al.,  SIK  (Salt-­‐inducible  kinase):  regulation  of  ACTH-­‐mediated  

steroidogenic  gene  expression  and  nuclear/cytosol  redistribution.  Endocr  Res,   2000.  26(4):  p.  995-­‐1002.  

41.   Jhala,  U.S.,  et  al.,  cAMP  promotes  pancreatic  beta-­‐cell  survival  via   CREB-­‐mediated  induction  of  IRS2.  Genes  Dev,  2003.  17(13):  p.  1575-­‐80.  

42.   Screaton,  R.A.,  et  al.,  The  CREB  coactivator  TORC2  functions  as  a  calcium-­‐  and   cAMP-­‐sensitive  coincidence  detector.  Cell,  2004.  119(1):  p.  61-­‐74.  

43.   Muraoka,  M.,  et  al.,  Involvement  of  SIK2/TORC2  signaling  cascade  in  the   regulation  of  insulin-­‐induced  PGC-­‐1alpha  and  UCP-­‐1  gene  expression  in  brown   adipocytes.  Am  J  Physiol  Endocrinol  Metab,  2009.  296(6):  p.  E1430-­‐9.  

44.   Park,  J.,  et  al.,  SIK2  is  critical  in  the  regulation  of  lipid  homeostasis  and   adipogenesis  in  vivo.  Diabetes,  2014.  63(11):  p.  3659-­‐73.  

45.   Sasaki,  T.,  et  al.,  SIK2  is  a  key  regulator  for  neuronal  survival  after  ischemia  via   TORC1-­‐CREB.  Neuron,  2011.  69(1):  p.  106-­‐19.  

46.   Horike,  N.,  et  al.,  Downregulation  of  SIK2  expression  promotes  the  melanogenic   program  in  mice.  Pigment  Cell  Melanoma  Res,  2010.  23(6):  p.  809-­‐19.  

47.   Ahmed,  A.A.,  et  al.,  SIK2  is  a  centrosome  kinase  required  for  bipolar  mitotic   spindle  formation  that  provides  a  potential  target  for  therapy  in  ovarian  cancer.  

Cancer  Cell,  2010.  18(2):  p.  109-­‐21.  

48.   Fu,  A.,  C.E.  Eberhard,  and  R.A.  Screaton,  Role  of  AMPK  in  pancreatic  beta  cell   function.  Mol  Cell  Endocrinol,  2013.  366(2):  p.  127-­‐34.  

49.   Sakamaki,  J.,  et  al.,  Role  of  the  SIK2-­‐p35-­‐PJA2  complex  in  pancreatic  beta-­‐cell   functional  compensation.  Nat  Cell  Biol,  2014.  16(3):  p.  234-­‐44.  

50.   Henriksson,  E.,  et  al.,  The  AMPK-­‐related  kinase  SIK2  is  regulated  by  cAMP  via   phosphorylation  at  Ser358  in  adipocytes.  Biochem  J,  2012.  444(3):  p.  503-­‐14.  

51.   Speidel,  D.,  et  al.,  CAPS1  regulates  catecholamine  loading  of  large  dense-­‐core   vesicles.  Neuron,  2005.  46(1):  p.  75-­‐88.  

52.   Cisternas,  F.A.,  et  al.,  Cloning  and  characterization  of  human  CADPS  and   CADPS2,  new  members  of  the  Ca2+-­‐dependent  activator  for  secretion  protein   family.  Genomics,  2003.  81(3):  p.  279-­‐91.  

53.   Khodthong,  C.,  et  al.,  Munc13  homology  domain-­‐1  in  CAPS/UNC31  mediates   SNARE  binding  required  for  priming  vesicle  exocytosis.  Cell  Metab,  2011.  14(2):  

p.  254-­‐63.  

54.   Nguyen  Truong,  C.Q.,  et  al.,  Secretory  vesicle  priming  by  CAPS  is  independent  of   its  SNARE-­‐binding  MUN  domain.  Cell  Rep,  2014.  9(3):  p.  902-­‐9.  

55.   Parsaud,  L.,  et  al.,  Calcium-­‐dependent  activator  protein  for  secretion  1  (CAPS1)   binds  to  syntaxin-­‐1  in  a  distinct  mode  from  Munc13-­‐1.  J  Biol  Chem,  2013.  

288(32):  p.  23050-­‐63.  

56.   Speidel,  D.,  et  al.,  A  family  of  Ca2+-­‐dependent  activator  proteins  for  secretion:  

comparative  analysis  of  structure,  expression,  localization,  and  function.  J  Biol   Chem,  2003.  278(52):  p.  52802-­‐9.  

57.   Binda,  A.V.,  N.  Kabbani,  and  R.  Levenson,  Regulation  of  dense  core  vesicle   release  from  PC12  cells  by  interaction  between  the  D2  dopamine  receptor  and   calcium-­‐dependent  activator  protein  for  secretion  (CAPS).  Biochem  Pharmacol,   2005.  69(10):  p.  1451-­‐61.  

58.   Shinoda,  Y.,  et  al.,  Calcium-­‐dependent  activator  protein  for  secretion  2  (CAPS2)   promotes  BDNF  secretion  and  is  critical  for  the  development  of  GABAergic   interneuron  network.  Proc  Natl  Acad  Sci  U  S  A,  2011.  108(1):  p.  373-­‐8.  

59.   Sadakata,  T.,  et  al.,  Autistic-­‐like  phenotypes  in  Cadps2-­‐knockout  mice  and   aberrant  CADPS2  splicing  in  autistic  patients.  J  Clin  Invest,  2007.  117(4):  p.  

931-­‐43.  

60.   Liu,  Y.,  et  al.,  CAPS  facilitates  filling  of  the  rapidly  releasable  pool  of  large   dense-­‐core  vesicles.  J  Neurosci,  2008.  28(21):  p.  5594-­‐601.  

61.   Neuman,  J.C.,  et  al.,  A  method  for  mouse  pancreatic  islet  isolation  and   intracellular  cAMP  determination.  J  Vis  Exp,  2014(88):  p.  e50374.  

62.   Berwin,  B.,  E.  Floor,  and  T.F.  Martin,  CAPS  (mammalian  UNC-­‐31)  protein  

localizes  to  membranes  involved  in  dense-­‐core  vesicle  exocytosis.  Neuron,  1998.  

21(1):  p.  137-­‐45.  

63.   Sadakata,  T.,  M.  Washida,  and  T.  Furuichi,  Alternative  splicing  variations  in   mouse  CAPS2:  differential  expression  and  functional  properties  of  splicing   variants.  BMC  Neurosci,  2007.  8:  p.  25.  

64.   Caromile,  L.A.,  et  al.,  The  neurosecretory  vesicle  protein  phogrin  functions  as  a   phosphatidylinositol  phosphatase  to  regulate  insulin  secretion.  J  Biol  Chem,  

2010.  285(14):  p.  10487-­‐96.  

Figures  

Figure 1. The expression SIK2 and CAPS2 in mouse pancreatic islets Immunohistochemistry of mouse pancreatic FFPE slides using SIK2, CAPS2, insulin antibodies. Alexa594 and Alexa488 secondary antibodies were used as negative control.

Scale bar = 25 µm.

(A)

(B)

(C)

Figure 2. Wholemount immunofluorescent staining of isolated islets

(A) Schematic of wholemount immunofluorescent staining procedures on isolated islets from ICR mice. (B) Bright field image of freshly isolated mouse islets in PBS. Scale bar

= 200 µm. (C) Wholemount immunofluorescent staining on isolated mouse islets with SIK2, CAPS2, insulin antibodies. Scale bar = 100 µm.

(A)

(B)

Figure 3. Subcellular localization of SIK2 and CAPS2 in mouse pancreatic β cell (A)Wholemount immunofluorescent staining on isolated mouse islets with SIK2, CAPS2, insulin antibodies by confocal imaging. Scale bar = 10 µm. (B) Z stack images of immunofluorescent staining on mouse pancreatic β cell co-stained with insulin and

CAPS2. Image inside dotted frame showing the co-localization of CAPS2 and insulin vesicles was magnified and presented with serial Z positions. Images were collected at 0.3 µm intervals. Scale bar = 10 µm and 2 µm.

(A)

(B)

Figure 4. Cloning of CAPS2 from mouse pancreatic islets

(A) Domain structure of seven identified mouse CAPS2 isoforms. Blue, green and yellow boxes represent C2 domain, PH domain, MHD domain respectively. Numbers refer to the amino acids defining each domains and dotted lines refer to alternatively spliced exons. (B) Construct map of pCMV-CAPS2 v7-WT cloned from cDNA of isolated mouse pancreatic islets. (C) Western blot analysis of HEK 293T cells

(C)

(D)

transfected with pCMV-3B vector (mock) or pCMV-CAPS2 v7-WT construct. The endogenous CAPS2 protein and exogenous myc-tagged CAPS2 recombinant protein were probed by CAPS2 (P-20, H-57) and myc antibodies. (D) Immunofluorescent staining of HEK 293T cells transfected with pCMV-3B vector (mock) or pCMV-CAPS2 v7-WT construct. The exogenous myc-tagged CAPS2 recombinant proteins were expressed and probbed by myc antibodies.

Figure 5. Physiological relevant proteins interaction between SIK2 and CAPS2 Co-immunoprecipitation of CAPS2 and SIK2 with different kinase activities. HEK 293T cells were co-transfected with pCMV-myc-CAPS2-WT construct as well as pCMV-flag-SIK2 constructs including Mock, SIK2-WT, SIK2-S587A, SIK2-S587D, SIK2- KD respectively. Flag-SIK2 recombinant proteins were immunoprecipitated by flag antibody and analyzed whether CAPS2 is its binging partner by western blot. A total of10%cell lysate was loaded in the input lanes.

(A)

(B)

(C)

Figure 6. Putative SIK2 phosphorylation sites on CAPS2 and CAPS2-MUT-T1016A, T1052A, T1231A construct

(A) Putative SIK2 phosphorylation sites on mouse CAPS2 predicted by Scansite in silico database mining (http://scansite.mit.edu) using SIK2 phosphorylation consensus motif sequence as search reference. SIK2 phosphorylation consensus motif sequence is (Hy){(B)X or X(B)}XX(S/T)XXX(Hy) (Hy: hydrophobic amino acid, B: basic amino acid, S/T: serine or threonine, X: any amino acid, respectively). Color code for the conserved amino acid residues: hydrophobic residues G, A, V, I, L, M (blue), basic residues H, K, R (green), S or T residues (red) and any other amino acids (black).

Threonine residues located at 1016, 1052, 1231 of mouse CAPS2 which scored 0.00 represents the most potential SIK2 phosphorylation sites. (B) Site direct mutagenesis of CAPS2 T1016, T1052, T1231 by the replacement of threonine with alanine. (C) Indicated primers for site direct mutagenesis.

Figure 7. In vitro phosphorylation of SIK2 on CAPS2

In vitro kinase assay using immunoprecipitates of Flag-SIK2 (Mock, WT, S587A, S587D, KD) as kinases and Myc-CAPS2 (WT or MUT) as substrates. Phosphorylation of CAPS2 was detected by p-Thr (phospho-threonine) antibody. Phosphorylation level of CAPS2 by SIK2 was measured as p-Thr CPAS2 / total CPAS2.

(A)

(B)

Figure 8. Transfection and expression of pTimer-phogrin in primarily cultured islet cells

(A) Immunofluorescent staining of primarily cultured β cells from isolated ICR mouse islets with insulin antibodies. Scale bar = 10 µm. (B) Images of primary β cells transfected with pTimer-phogrin and detected by 488 nm laser. Immunofluorescent staining with insulin antibodies showed phogrin was co-localized with insulin vesicles (white arrows). Scale bar = 5 µm.

(A)

(B)

(C)

(E)

(D)

Figure 9. Insulin secretion of SIK2/CAPS2 overexpressed mouse primary β cells (A) Schematic illustration of experimental design. Primarily cultured islets were transfected with Mock, CAPS2, SIK2, and SIK2 plus CAPS2 according to modified protocol. 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. Secreted insulin were measured by mouse insulin ELISA. (B) Respectively secreted insulin from islets transfected with Mock (n=4), CAPS2 (n=2), SIK2 (n=3), and SIK2 plus CAPS2 (n=3). (C) Average of secreted insulin from each group. (D) The glucose stimulated insulin secretion from 0 to 15 min. (E)The proposed model showing how SIK2 regulates insulin vesicle release upon glucose stimulation.

Only if both SIK2 and CAPS2 are overexpressed, the amount SIK2/CAPS2 complexes increase. As the consequence, the rapid insulin release is promoted.

(A)

(B) s

Figure 10. Proposed model of the SIK2/CAPS2 pathway

(A) Condition without glucose stimulation. SIK2 with active kinase activity would phosphorylate CAPS2, thereby interfere the binding between two proteins. (B) Condition with glucose stimulation. The cAMP-dependant PKA phosphorylation of SIK2-S587 inhibits its kinase activity. Unphosphorylated CAPS2 is able to form SIK2-CAPS2 complex which promotes insulin vesicle release.

Appendix

(A)

(B)

Appendix 1. Dominant expression of SIK2 rather than SIK1 or SIK3 in pancreatic islets mainly in insulin producing β cells

(A) Immunohistochemistry of FFPE pancreatic tissue slids from SD rat at 8 week of age with insulin, SIK1, SIK2 and SIK3 antibodies (red arrow). (B) Immunofluorescent staining of FFPE mouse pancreatic tissue slids with insulin and SIK2 antibodies.

(Data from Dr. Han-Yi Chou’s lab: Pei-Han Tai)

(A)

(B)

Appendix 2. Co-fractionation of SIK2 with insulin-containing LDCV in Rinm5F cells

(A) Schematic of biochemical sucrose gradient fractionation procedures of insulin vesicles form Rinm5F insulinoma cells (B) Insulin concentration of each fraction measured by insulin ELISA and fractionated samples precipitated with TCA (trichloroacetic acid) and analyzed by Western blot for the indicated proteins.

(Data from Dr. Han-Yi Chou’s lab: I-I Chuo)

(A)

(B)

(D)

Appendix 3. Phosphorylation of SIK2-S587 by cAMP dependant PKA in β cells upon glucose stimulation (A) Immunofluorescent staining of RINm5F insulinoma cells with insulin and SIK2-S587 antibodies. SIK2-S587 was co-localized with insulin vesicles. (B) Co-immunoprecipitation of SIK2 and PKAβ. PKAβ instead of PKAα was presented in the same protein complex containing SIK2. (C) In vitro kinase assay of SIK2-WT and SIK2-S587A with the treatment of PKA activator (FSK) or PKA inhibitor. Phosphorylation level of SIK2-S587 was detected by SIK2-S587 and SIK2 antibodies. (D)(E) Western blot analysis of isolated mouse islets in starvation or treated with 16.7mM glucose or compound C (SIK2 inhibitor). SIK2 was phosphorylated at Ser587 under glucose or compound C treatments.

(Data from Dr. Han-Yi Chou’s lab: Pei-Han Tai and Chung-Ju Kao) (C)

(E)

(A)

(B)

Appendix 4. Involvement of SIK2 kinase activity in insulin secretion

(A) Immunofluorescent staining of RINm5F cells with SIK2, SIK2-S587, SIK2-T175 and insulin antibodies. The kinase active SIK2-T175 and kinase inactive SIK2-S587

(A) Immunofluorescent staining of RINm5F cells with SIK2, SIK2-S587, SIK2-T175 and insulin antibodies. The kinase active SIK2-T175 and kinase inactive SIK2-S587

相關文件