Abstract

   Islet α-cell dysfunction has been shown to contribute to type 2
   diabetes; however, whether islet α-cell inflammation is involved in the
   occurrence of pancreatitis is largely unknown. The aims of this study
   were to investigate how NF-κB inducing kinase (NIK) regulates
   pancreatic α-cell function, both in vitro and in vivo, and to assess
   how islet α-cell inflammation induced by NIK affects the development of
   pancreatitis.

   Methods: We utilized adenovirus-mediated NIK overexpression, ELISA,
   qPCR, RNA-seq, and Western blot analyses to study the role of NIK in
   islet α cells in vitro. Islet α-cell-specific NIK overexpressing
   (α-NIK-OE) mice were generated, and pancreatic α/β-cell function and
   the occurrence of pancreatitis in these mice were assessed via ELISA,
   qPCR, and immunohistochemical analyses.

   Results: The LTβR/noncanonical NF-κB signaling pathway is present in
   islet α cells. Overexpression of NIK in αTC1-6 cells induces
   inflammation and cell death, contributing to a decrease in the
   expression and secretion of glucagon. Additionally, α-cell specific
   overexpression of NIK (α-NIK-OE) results in α-cell death, lower serum
   glucagon levels, and hypoglycemia in mice. Strikingly, α-NIK-OE mice
   also display a reduced β-cell mass, growth retardation, pancreatitis,
   and postnatal death.

   Conclusions: Islet α-cell specific overexpression of NIK results in
   islet α-cell dysfunction and causes islet β-cell death and
   pancreatitis, which are most likely due to paracrine secretion of
   cytokines and chemokines from islet α cells, thus leading to
   hypoglycemia, growth retardation, and postnatal death in mice.

   Keywords: islet α cell, NF-κB inducing kinase, inflammation,
   pancreatitis, chemokine

Introduction

   The pancreas contains endocrine islets and exocrine acinar cells, which
   produce hormones and digestive enzymes that are key regulators of
   glucose metabolism and food digestion. Islet β-cell damage can lead to
   diabetes, while injury of acinar cells can cause pancreatitis, with
   inflammation being one of primary factors involve in the development of
   both diabetes and pancreatitis [49]^1^, [50]^2. Islet β-cell
   inflammation has been shown to contribute to pancreatitis [51]^3, but
   whether or not islet α-cell inflammation can lead to pancreatitis is
   largely unknown.

   NF-κB-inducing kinase (NIK) is a widely expressed Ser/Thr kinase that
   functions as a central mediator of noncanonical NF-κB activation
   [52]^4; however, its expression is quite low in quiescent cells due to
   TRAF2/TRAF3/cIAP1/2-mediated ubiquitination and degradation [53]^4. In
   the presence of specific cytokines, such as BAFF, CD40L, LTβ, and
   RANKL, receptors for these cytokines are activated, which promotes the
   degradation of TRAF2/TRAF3/cIAP1/ 2, leading to the stabilization and
   activation of NIK. In turn, NIK phosphorylates and activates IKK α
   [54]^4^, [55]^5, and activated IKKα phosphorylates p100 (the inactive
   full-length NF-κB2), promoting its proteolytic cleavage to produce p52
   (the active form of NF-κB2) [56]^6. Several studies have shown that NIK
   can regulate T/B-cell development [57]^7^, [58]^8, and influence liver
   fibrosis [59]^9^, [60]^10, and metabolic syndrome in mice [61]^11;
   however, the role of NIK in islet α cells is not completely understood.
   Whether islet α-cell inflammation induced by NIK can contribute to
   pancreatitis is also largely unknown.

   In this study, we demonstrate that overexpression of NIK in islet α
   cells decreases glucagon secretion and induces inflammation and cell
   death both in vitro and in vivo. This further promotes the cell death
   of islet β cells and pancreatic acinar cells via increased paracrine
   secretion of inflammatory cytokines and chemokines, thus leading to
   hypoglycemia, growth retardation, pancreatitis, and postnatal death in
   mice.

Materials and Methods

Animal experiments

   Animal experiments were carried out in strict accordance with the Guide
   for the Care and Use of Laboratory Animals of the National Institutes
   of Health and were approved by the Animal Experimental Ethics Committee
   of Northeast Normal University and Harbin Institute of Technology.
   STOP-NIK and Glucagon-cre mice (C57BL/6 background) were described
   previously [62]^9^, [63]^10^, [64]^12^, [65]^13. STOP-NIK mice were
   crossed with glucagon-cre mice to generate islet α-cell-specific NIK
   overexpression (α-NIK-OE: STOP-NIK^+/-; Glucagon- Cre^+/-) mice.
   Control mice were their littermates (Genotype: STOP-NIK^+/-).
   ROSA26-EYFP reporter mice were purchased from Shanghai Biomodel
   Organism Science & Technology Development Co., Ltd. Mice were housed on
   a 12-h light/12-h dark cycle, and were fed with a normal chow and free
   access to water. Male littermates were used for experiments. Blood
   glucose levels were measured as desscribed previsouly [66]^10. Blood
   samples were collected from orbital sinus. Serum glucagon and insulin
   levels were measured using glucagon ELISA kits (DGCG0, R&D Systems) and
   insulin ELISA kits (EZRMI-13K, Millipore Corporation), respectively.
   Serum amylase activity was measured using α-Amylase assay kits (C016-1,
   Nanjing Jiancheng Bioengineering Institute). Pancreatic trypsin
   activity was measured using Trypsin ELISA kits ([67]D59091,
   Immuno-Biological Laboratories Co., Ltd.) following the manufacturer's
   recommended procedure.

   For cerulein-induced acute pancreatitis, 9-week old male C57BL/6 mice
   were intraperitoneally injected with 50 μg/kg cerulein (Sigma-Aldrich,
   St. Louis, MO) in saline every hour for a total of seven injections.
   Mice were sacrificed at 12 h time point, and pancreases were fixed with
   4% paraformaldehyde and subjected to immunostaining assays.

Pancreatic islet and acinar cell isolation

   Male mice were euthanized. Pancreases were cut into small pieces, and
   digested with 1 mg/mL collagenase P (Roche Diagnostics) in Hanks'
   balanced salt solution (HBSS) as shown previously [68]^14. Pancreatic
   islets and acinar cells were hand-picked.

Transient transfection and luciferase assays

   HEK293 cells were divided equally in a 24-well plate and cultured
   overnight. The cells were cotransfected with mouse glucagon promoter
   (-1000-0 bp) luciferase reporter plasmid with NIK or p52 at different
   doses (0, 100, 200, 400 ng) for 24 h. The cells were then harvested in
   reporter lysis buffer (Promega, Madison, WI, USA). Luciferase activity
   was measured and normalized to β-Gal activity as shown previously
   [69]^10.

Cell culture, adenoviral infection, and low glucose-stimulated glucagon
secretion (LGSGS)

   αTC1-6 cells (a mouse pancreatic alpha cell line) were cultured at 37°C
   in 5% CO[2] in DMEM supplemented with 100 units ml^-1 penicillin, 100
   units ml^-1 streptomycin, and 10% FBS. INS-1 832/13 cells (a rat
   insulinoma cell line) were cultured at 37°C and 5% CO[2] in RPMI-1640
   medium supplemented with 10% FBS and 50 mM β-mercaptoethanol as shown
   previously [70]^15^, [71]^16. β-Gal, NIK, and NIK(KA) adenovirus were
   described before [72]^10^, [73]^17. αTC1-6 cells were infected with
   β-Gal and NIK adenovirus for 48 h and subjected to MTT and TUNEL
   assays. For LGSGS assay, αTC1-6 cells were infected with β-Gal and NIK
   adenovirus for 16 h, and these cells were incubated at 37°C in 200 μL
   of HBSS (pH 7.4) containing 25 mM or 1 mM glucose for 1 h. Medium was
   collected to measure LGSGS. Cells were then harvested in a lysis
   buffer, and protein concentrations were measured. The cell extracts
   were then mixed with acid-ethanol (1.5% HCl in 70% EtOH) and were used
   to measure glucagon content. Glucagon secretion was normalized to
   protein levels.

Immunoblotting

   αTC1-6 cells were harvested in a lysis buffer (50 mM Tris HCl, pH 7.5,
   1.0% NP-40, 150 mM NaCl, 2 mM EGTA, 1 mM Na[3]VO[4], 100 mM NaF, 10 mM
   Na[4]P[2]O[7], 1 mM PMSF, 10 μg/mL aprotinin, 10 μg/mL leupeptin). Cell
   extracts were immunoblotted with the indicated antibodies and were
   visualized using the ECL. Antibody dilution ratios were as follows:
   Flag (F1804, Sigma, 1:5000 dilution), NF-κB2 (4882, Cell Signaling
   Technology, 1:2000 dilution), Tubulin (sc5286, Santa Cruz, 1:5000
   dilution).

Quantitative real-time PCR (qPCR) analysis

   αTC1-6 cells were infected with β-Gal, NIK and NIK(KA) adenovirus for
   16 h. Total RNAs were extracted using TriPure Isolation Reagent (Roche,
   Mannheim, Germany), and the first-strand cDNAs were synthesized using
   random primers and M-MLV reverse transcriptase (Promega, Madison, WI)
   as shown before [74]^10. RNA abundance was measured using ABsolute qPCR
   SYBR Mix (Roche, Mannheim, Germany) and Roche LightCycler 480 real-time
   PCR system (Roche, Mannheim, Germany). The expression of individual
   genes was normalized to the expression of 36B4, a house-keeping gene.
   Primers for real time qPCR were listed in Table [75]S1.

Treatment of INS-1 832/13 cells with conditioned medium from αTC1-6 cell

   αTC1-6 cells were infected with β-Gal and NIK adenovirus for 18 h, and
   then the cells were grown in fresh serum-free DMEM for 24 h.
   Conditioned media from NIK-overexpressing αTC1-6 cells (NIKCM) and
   control cells (βGalCM) were then collected. Half of NIKCM was
   neutralized with antibodies (Proteintech) against TNFα, CCL2 and CCL5,
   denoted as NIKCM+Ab. INS-1 832/13 cells were treated with these
   conditioned media for 16 h. MTT and TUNEL assays were performed as
   shown previously [76]^15. TNFα, CCL2, and CCL5 protein levels in these
   conditioned media and αTC1-6 cells were measured by ELISA kits
   (Proteintech). For signal detection, Streptavidin-HRP was added,
   followed by tetramethyl-benzidine (TMB) reagent. Solution containing
   sulfuric acid was used to stop color development and the color
   intensity which was proportional to the quantity of bound protein was
   measurable at 450 nm.

Immunostaining and TUNEL assays

   Pancreata were fixed in 4% paraformaldehyde for 3 h and then in 30%
   sucrose overnight. Frozen pancreatic sections (5-8 μm) were stained
   with the indicated antibodies. For the measurement of islet glucagon
   and insulin positive area, pancreatic sections were immunostained with
   anti-glucagon and anti-insulin antibodies. Islet glucagon and insulin
   positive areas were measured using Image J software, and normalized to
   total pancreatic section area as described before [77]^15. Antibody
   dilution ratios were as follows: insulin (A0564, Dako, 1:1000
   dilution); glucagon (G2654, Sigma, 1:1000 dilution); and Ki67 (VP-RM04,
   Vector Laboratories, 1:100 dilution); F4/80 (14-4801-82, eBiosience,
   1:200 dilution); LTβR (20331-1-AP, Proteintech, 1:200 dilution); NF-κB2
   (4882, Cell Signaling Technology, 1:100 dilution); CCL2 (66272-1-Ig,
   Proteintech, 1:100 dilution). Frozen pancreatic sections, αTC1-6, and
   INS-1 832/13 cells were fixed with 4% paraformaldehyde and subjected to
   TUNEL assays using cell death detection kits (Roche Diagnostics)
   following the manufacturer's recommended procedure. The samples were
   stained with DAPI to visualize total cells. Pancreatic sections were
   also immunostained with anti-glucagon antibody to identify islet α
   cells or with anti-insulin antibody to identify islet β cells.
   TUNEL-positive cells were counted and normalized to total islet cell
   number.

RNA-seq

   αTC1-6 cells were infected with β-Gal, NIK and NIK(KA) adenovirus for
   16 h. Total RNAs were extracted using TriPure Isolation Reagent (Roche,
   Mannheim, Germany). mRNA profiles of β-Gal-, NIK-, and
   NIK(KA)-overexpressing αTC1-6 cells were generated by deep sequencing,
   in a single test, using Illumina. RNA sequencing libraries were
   generated using NEBNext® Ultra™ RNA Library Prep Kit for Illumina®
   (NEB, USA) following manufacturer's recommendations and library quality
   was assessed on the Agilent Bioanalyzer 2100 system. The mRNA library
   was sequenced on an Illumina Hiseq platform and 125 bp/150 bp
   paired-end reads were generated. Paired-end clean reads were aligned to
   the mouse reference genome (Ensemble_GRCm38.89) with TopHat (version
   2.0.12), and the aligned reads were used to quantify mRNA expression by
   using HTSeq-count (version 0.6.1).Differential expression analysis of
   two conditions was performed using the DEGSeq R package (1.20.0). The P
   values were adjusted using the Benjamini & Hochberg method. Corrected
   P-value of 0.005 and log2 (Fold change) of 1 were set as the threshold
   for significantly differential expression. Gene Ontology (GO)
   enrichment analysis of differentially expressed genes was implemented
   by the GOseq R package, in which gene length bias was corrected. GO
   terms with corrected P value less than 0.05 were considered
   significantly enriched by differential expressed genes. KEGG is a
   database resource for understanding high-level functions and utilities
   of the biological system, such as the cell, the organism and the
   ecosystem, from molecular-level information, especially large-scale
   molecular datasets generated by genome sequencing and other
   high-through put experimental technologies
   ([78]http://www.genome.jp/kegg/). We used KOBAS software to test the
   statistical enrichment of differential expression genes in KEGG
   pathways. RNA-seq data that support the findings of this study have
   been deposited in GEO under accession code [79]GSE112511.

Statistical Analysis

   Data were presented as mean ±SEM. Differences between groups were
   analyzed by two-tailed Student's t tests. P< 0.05 was considered
   statistically significant.

Results

The LTβR/noncanonical NF-κB signaling pathway is present in pancreatic islet
α cells

   It has been shown that noncanonical NF-κB signaling can be stimulated
   by the activation of BAFF-R, CD40, LTβR, or RANK. To determine the
   relative expression levels of these four receptors, we performed qPCR
   assays in αTC1-6 cells (a mouse pancreatic α cell line). As shown in
   Figure [80]1A, LTβR was highly expressed in αTC1-6 cells; however,
   BAFF-R, CD40, and RANK were barely detectable (Figure [81]1A). Compared
   with pancreatic acinar cells, αTC1-6 cells and pancreatic islets
   displayed higher levels of LTβR (Figure [82]1B and [83]1C). Consistent
   with these observations, immunofluorescence data indicated that LTβR
   was highly expressed in pancreatic islet α cells (Figure [84]1C). We
   noticed that insulin positive cells also displayed high levels of LTβR
   (Figure [85]S1). Additionally, treatment with an anti-LTβR antibody
   resulted in increased p52 protein levels in αTC1-6 cells (Figure
   [86]1D). LIGHT, a ligand for LTβR, was able to increase p52 protein
   level (Figure [87]2A) and suppress low glucose-stimulated glucagon
   secretion (LGSGS) in αTC1-6 cells (Figure [88]2B). LIGHT also induced
   α-cell death (Figure [89]2C). These data demonstrated that the
   components of the LTβR/noncanonical NF-κB signaling pathway are present
   in pancreatic α cells, which may regulate islet α cell function.

Figure 1.

   [90]Figure 1
   [91]Open in a new tab

   LTβR/noncanonicalNF-κB2 signaling pathway is present in islet α cells.
   A, The relative expression of BAFF-R, CD40, LTβR, and RANK were
   measured by qPCR (left) and semi-quantitative PCR (right) in αTC1-6
   cells (n = 6/group). B, The relative expression of LTβR was measured by
   qPCR in αTC1-6 cells, pancreatic acinar cells, and pancreatic islets (n
   = 6/group). C, Immunostaining of LTβR and glucagon in the section of
   pancreas. D, αTC1-6 cells were treated with anti-LTβR antibody at
   different concentrations (0, 2, 10 μg/mL) for 2 h, and p52 protein
   levels were measured by Western blot. **, p < 0.01.

Figure 2.

   [92]Figure 2
   [93]Open in a new tab

   LIGHT activates noncanonical NF-κB2 signaling pathway, impairs glucagon
   secretion and increases cell death in αTC1-6 cells. A, αTC1-6 cells
   were treated with LIGHT at 0 and 10 ng/mL for 2 h. p100 and p52 protein
   levels were measured by Western blot. B-C, αTC1-6 cells were treated
   with LIGHT at 0 and 10 ng/mL for 16 h. Low glucose (1 mM) stimulated
   glucagon secretion (LGSGS) was measured by ELISA (B), and
   TUNEL-positive cells were measured (C). n = 3-4. **, p < 0.01.

NIK overexpression impairs glucagon secretion and increases cell death in
αTC1-6 cells

   NIK is the key activator of the noncanonical NF-κB signaling pathway
   [94]^4. To determine whether NIK can regulate islet α cell function, we
   overexpressed NIK in αTC1-6 cells by infecting them with an ad-NIK
   adenovirus and then assessed glucagon secretion and cell viability. NIK
   and p52 protein levels were very high in ad-NIK infected αTC1-6 cells
   (Figure [95]3A). As shown in Figure [96]3B, LGSGS was present in the
   β-Gal control group, which was completely blocked in NIK overexpressing
   αTC1-6 cells. Moreover, glucagon mRNA and protein levels were
   significantly decreased in αTC1-6 cells after activation of NIK
   signaling (Figure [97]3C and [98]3D). NIK and p52 dose-dependently
   inhibited glucagon promoter luciferase activity (Figure [99]3E and
   [100]3F). These data suggested that NIK/p52 directly inhibited glucagon
   gene transcription. Additionally, cell viability was significantly
   decreased and the number of TUNEL-positive cells was dramatically
   increased in NIK-overexpressing αTC1-6 cells (Figure [101]3G and
   [102]3H), which also contributed to the observed impairment in LGSGS
   and reduced glucagon expression.

Figure 3.

   [103]Figure 3
   [104]Open in a new tab

   NIK overexpression impairs glucagon secretion and increases cell death
   in αTC1-6 cells. αTC1-6 cells were infected with β-Gal and NIK
   adenovirus for 16 h. A, NIK, p52, and Tubulin protein levels were
   measured by Western blot. B, Low glucose (1 mM) stimulated glucagon
   secretion (LGSGS) was measured by ELISA (n = 3-4). C, Glucagon mRNA was
   measured by RT-qPCR (n = 5). D, Glucagon content was measured by ELISA
   (n = 4). E-F, Glucagon promoter luciferase reporter and NIK/p52
   plasmids were cotransfected into HEK293 cells for 24 h. Luciferase
   activity was then measured and normalized to β-Gal activity (n = 6).
   G-H, Cell viability (MTT) and the number of TUNEL-positive cells were
   measured after virus infection for 48 h (n = 5-6). *, p< 0.05. **, p <
   0.01.

NIK overexpression triggers an immune response, inflammation, and chemokine
secretion in αTC1-6 cells

   To further determine the molecular mechanisms involved in the NIK
   regulation of α cell function, we performed RNA-seq analysis in
   β-Gal-overexpressing (control), NIK-overexpressing, and NIK(KA) (the
   dominate negative form of NIK)-overexpressing αTC1-6 cells. As shown in
   Figure [105]4A and [106]4D, NIK(KA)-overexpressing αTC1-6 cells
   displayed a very similar gene expression pattern with β-Gal control.
   However, 328 genes were upregulated and 48 genes were downregulated
   following NIK overexpression (Figure [107]4B). GO analysis indicated
   that the upregulated genes were primarily related to the immune
   response (Figure [108]4C). KEGG pathway enrichment analysis showed that
   immune signaling pathways, including the TNF, NF-κB, and chemokine
   signaling pathways, were significantly activated by NIK overexpression
   in αTC1-6 cells (Figure [109]S2A), while genes related to
   metabolism-associated signaling pathways were significantly decreased
   (Figure [110]S2B). Among the upregulated genes, the expression of the
   highest-ranked cytokines and chemokines were confirmed by qPCR (Figure
   [111]4D) and ELISA (Figure [112]4E). CXCL1 and iNOS mRNA levels were
   increased by more than 1500-fold, and CCL2, CXCL2, and CXCL10 mRNA
   levels were increased by more than 400-fold (Figure [113]4D). TNFα,
   CCL5, CXCL3, CXCL5, and CSF1 mRNA levels were also significantly
   increased (Figure [114]4D). Moreover, results of ELISA clearly showed
   that TNFα, CCL2, and CCL5 were secreted from αTC1-6 cells infected with
   ad-NIK (Figure [115]4F). These data indicated that NIK overexpression
   impairs glucagon secretion and induces cell death due to the activation
   of an immune response and inflammation in αTC1-6 cells.

Figure 4.

   [116]Figure 4
   [117]Open in a new tab

   NIK overexpression triggers an immune response, inflammation, and
   chemokine secretion in αTC1-6 cells. αTC1-6 cells were infected with
   β-Gal, NIK and NIK(KA) adenovirus for 16 h. RNA-seq analysis in both
   β-Gal-overexpressing (control), NIK-overexpressing, and
   NIK(KA)-overexpressing αTC1-6 cells. A, Heat map derived from RNA-Seq
   expression data ofthe differentially expressed genes (DEGs). B, DEGs
   (NIK VS β-Gal). C, Top GO biological process terms enriched in
   upregulated genes in NIK-overexpressing αTC1-6 cells compared with
   β-Gal control. D, Relative TNFα, CCL2, CCL5, CSCL1, CXCL2, CX3CL1,
   CXCL5, CXCL10, CSF1, and iNOS mRNA levels were measured by RT-qPCR (n =
   6). E-F, TNFα, CCL2, and CCL5 protein levels in cell lysis and
   supernatant were measured by ELISA (n = 5-8). *, p < 0.05. **, p <
   0.01.

Islet α-cell specific overexpression of NIK (α-NIK-OE) causes postnatal
death, lower body weight, hypoglycemia, and impaired islet structure and
function in mice

   To examine the role of NIK in pancreatic α cells in vivo, we generated
   α-cell specific NIK overexpression (α-NIK-OE) mice by crossing STOP-NIK
   mice with glucagon-cre transgenic mice. The specificity of the
   glucagon-promoter driven Cre recombinase (glucagon-cre) was confirmed
   by crossing ROSA26-EYFP reporter mice with glucagon-cre mice and
   immunostaining for EYFP in pancreatic islets, hypothalamus, and
   intestine cells. As shown in Figure [118]5A, more than 95% of
   glucagon-positive cells expressed EYFP and none of insulin positive
   cells expressed EYFP, suggesting a high specificity of glucagon-cre
   expressed in islet α cells. We noticed that very few (about 0.55%)
   YFP-positive cells expressed somatostatin (Figure [119]5A). We also
   observed YFP signal in a very small amount of intestine L cells and
   neurons in the brain, which is consistent with previous observations
   [120]^13.

Figure 5.

   [121]Figure 5
   [122]Open in a new tab

   Islet α-cell-specific overexpression of NIK causes postnatal death,
   lower body weight, hypoglycemia, and impaired islet structure and
   function in mice. A, Co-immunostaining of YFP and glucagon or YFP and
   insulin in pancreatic sections of Glucagon-cre/ROSA26-YFP mice. B,
   Pancreatic islets were isolated from α-NIK-OE mice and control
   littermates at 27 days age. NIK mRNA levels were measured by RT-qPCR
   and normalized to 36B4 levels (n = 4/group). C, Representative α-NIK-OE
   mice and control littermates at 23 days age. D, Growth curve of
   α-NIK-OE mice and control littermates (n = 7-14/group). E, Survival
   curves (n = 29/group). F, Fed blood glucose levels (n = 9-14/group).
   G-H, Serum glucagon and insulin levels at 20-28 days age (n =
   10-15/group). I-J, Pancreas glucagon and insulin contents at 20-28 days
   age (n = 9-12/group). K, Islet number/section (n = 6/group). L,
   Representative islet from α-NIK-OE mice and control littermates was
   immunostained with insulin and glucagon. M-N, Relative glucagon- and
   insulin-positive areas. O, Ratio of islet glucagon /insulin-positive
   areas in α-NIK-OE mice and control littermates (n = 5-6/group).*, p <
   0.05. **, p < 0.01.

   NIK mRNA levels were increased by ∼8-fold in pancreatic islets of
   α-NIK-OE mice (Figure [123]5B). α-NIK-OE mice did not gain any body
   weight (Figure [124]5C and [125]5D), and died within 40 days after
   birth (Figure [126]5E), and the fed blood glucose levels in α-NIK-OE
   mice gradually decreased after birth (Figure [127]5F). As glucagon and
   insulin are two key hormones that tightly control blood glucose levels,
   we measured the serum glucagon and insulin levels in α-NIK-OE and
   control mice. As shown in Figure [128]5G, the serum glucagon levels
   were significantly reduced by 55.6% in α-NIK-OE mice. Surprisingly, the
   serum insulin levels were also dramatically decreased in these mice
   (Figure [129]5H). Accordingly, pancreatic glucagon and insulin contents
   were significantly decreased in α-NIK-OE mice relative to the control
   mice (Figure [130]5I and 5J).

   To determine whether the reduced glucagon and insulin levels in
   α-NIK-OE mice were caused by a reduction in islet insulin- and
   glucagon-positive areas, pancreatic sections from α-NIK-OE and control
   mice were stained using antibodies against insulin and glucagon.
   Compared with the control mice, islet number, the size of individual
   islet and total islet area were significantly decreased in α-NIK-OE
   mice (Figure [131]5K and 5L). The relative glucagon-positive area was
   also lower in α-NIK-OE mice (Figure [132]5L and 5M), and the relative
   insulin-positive area was also lower in these mice (Figure [133]5L and
   5N). The ratio of islet glucagon/insulin-positive area was
   significantly reduced in α-NIK-OE mice (Figure [134]5O), indicating
   that the reduction of islet α cell area was substantially more than
   that in the control mice.

   To determine whether cell death contributed to the observed reduction
   in glucagon- or insulin-positive areas in α-NIK-OE mice, we measured
   islet α- and β-cell apoptosis via TUNEL staining. Pancreatic sections
   were co-immunostained with anti-glucagon and anti-insulin antibodies to
   visualize α and β cells, respectively. The number of TUNEL-positive α
   or β cells was substantially more in α-NIK-OE mice than that in control
   mice (Figure [135]6A and [136]6B). We also examined islet α- and β-cell
   proliferation by immunostaining pancreatic sections with an anti-Ki67
   antibody, and islets from α-NIK-OE mice demonstrated markedly fewer
   Ki67-positive α and β cells (Figure [137]6C and [138]6D). Together,
   these data suggested that increased apoptosis and decreased
   proliferation of α and β cells contributed to the reduced islet mass
   and the decreased serum glucagon and insulin levels observed in
   α-NIK-OE mice.

Figure 6.

   [139]Figure 6
   [140]Open in a new tab

   Islet α-cell-specific overexpression of NIK increases cell death and
   decreased cell proliferation in both islet α and β cells. A-B, Islet α-
   and β-cell apoptosis in α-NIK-OE mice and control littermates at 17-27
   days age was measured by TUNEL assays. C-D, Islet α- and β-cell
   proliferation was measured by ki67 staining. Pancreatic sections were
   also immunostained with anti-glucagon antibody to identify islet α
   cells or with anti-insulin antibody to identify islet β cells (n =
   5-6/group). *, p < 0.05. **, p < 0.01.

   α-Cell-specific overexpression of NIK induces both α-cell death and
   β-cell death, indicating a crosstalk between α and β cells. To test
   this hypothesis, we collected conditioned media from NIK-overexpressing
   αTC1-6 cells (NIKCM) and control cells (βGalCM), and then treated INS-1
   832/13 cells with these conditioned media. As shown in Figure [141]S3A,
   cell viability was significantly decreased in INS-1 832/13 cells
   treated with NIKCM compared with the βGalCM group. Consistent with
   these observations, the presence of TUNEL-positive cells was
   dramatically increased in NIKCM-treated INS-1 832/13 cells (Figure
   [142]S3B), which was most likely due to high levels of TNFα, CCL2, and
   CCL5 in NIKCM (Figure [143]4F). A combination of TNFα, CCL2 and CCL5
   was able to induce β-cell death (Figure [144]S3C). After neutralization
   with antibodies against TNFα, CCL2 and CCL5, NIKCM was unable to induce
   β-cell death (Figure [145]S3A and S3B). These data suggested that
   activation of NIK in α cells resulted in β-cell death via paracrine
   secretion of cytokines and chemokines.

   In addition to the role of islets in the regulation of blood glucose
   levels, the liver is sensitive to increased glucagon levels secreted by
   islet α cells in response to hypoglycemia. To address the cause of the
   hypoglycemia in α-NIK-OE mice, we measured the expression of the
   gluconeogenesis related genes G6pase and PEPCK using qPCR. The
   expression levels of G6pase and PEPCK were relatively high in α-NIK-OE
   mice as compared to those in control mice (Figure [146]S4A), which may
   be a compensatory response to the hypoglycemia and lower serum glucagon
   levels present in α-NIK-OE mice. Importantly, normal liver structure
   was observed in α-NIK-OE mice (Figure [147]S4B). These data indicated
   that lower serum glucagon levels could be the primary cause of
   hypoglycemia in α-NIK-OE mice.

α-Cell-specific overexpression of NIK induces pancreatitis

   Several reports have shown that glucagon deficiency in mice only
   slightly decreases blood glucose levels without affecting growth or
   postnatal survival [148]^18^, [149]^19, which indicates that other
   factors in addition to islet dysfunction contribute to the growth
   retardation and postnatal death observed in α-NIK-OE mice. Nuclear
   NF-κB2 is increased in pancreatic islet α cells in acute pancreatitis
   mouse model (Figure [150]S5), indicating that NIK/p52 activation in
   islet α cells may be associated with pancreatitis. To assess the
   function of the pancreas, the size of the pancreas was first compared
   between α-NIK-OE and control mice. The pancreas-to-body-weight ratio
   was significantly decreased by 31.7% in α-NIK-OE mice relative to the
   control mice (Figure [151]7A). To check whether the smaller pancreas in
   these mice was due to increased apoptosis and/or decreased cell
   proliferation of acinar cells, pancreas sections were stained with
   TUNEL and Ki67. Figure [152]7B and [153]7C clearly showed that the
   number of TUNEL-positive cells was significantly increased while the
   presence of Ki67-positive cells was dramatically reduced in α-NIK-OE
   mice. We further checked the presence of pancreatitis markers in
   α-NIK-OE mice. The level of amylase was significantly increased in
   α-NIK-OE mice (Figure [154]7D), indicating tissue damage within the
   pancreas. Activation of trypsin is considered a key event in the onset
   of pancreatitis; as such, we next measured trypsin activity in the
   pancreas of α-NIK-OE mice. Figure [155]7E clearly showed that trypsin
   activity was significantly increased in α-NIK-OE mice. Moreover, H&E
   staining of pancreas sections showed that α-NIK-OE mice developed
   significantly more edema (Figure [156]7F). The relative F4/80-positive
   area was dramatically increased in the pancreas of α-NIK-OE mice
   (Figure [157]7G), indicating the infiltration of inflammatory cells.
   Moreover, the expression of inflammatory cytokine (TNFα) and chemokines
   (CCL2, CCL5, CXCL2, and CX3CL1) was dramatically increased in α-NIK-OE
   mice (Figure [158]7H). Immunostaining data showed that CCL2 colocalized
   with F4/80 (Figure [159]7G), indicating that increased chemokines were
   most likely from infiltrated F4/80-positive cells. Furthermore, CCL2
   and F4/80 were also localized in islets of α-NIK-OE mice (Figure
   [160]S6A and S6B), suggesting that α-NIK-OE mice also developed
   insulitis. Consistent with these observations, serum CCL2 and CCL5
   levels were also increased in α-NIK-OE mice (Figure [161]7I), and a
   combination of TNFα, CCL2 and CCL5 was able to induce acinar cell death
   (Figure [162]S7). However, we did not observe any obvious inflammation
   in the small intestines of α-NIK-OE mice (Figure [163]S8). Taken
   together, these data demonstrated that islet α-cell-specific
   overexpression of NIK induces islet cell dysfunction and pancreatitis,
   leading to hypoglycemia, growth retardation, and postnatal death in
   mice.

Figure 7.

   [164]Figure 7
   [165]Open in a new tab

   Islet α-cell-specific overexpression of NIK induces pancreatitis. The
   pancreas of α-NIK-OE mice and control littermates at 20-28 days age
   were analyzed. A, Pancreas weight/body weight (n = 9/group). B-C, TUNEL
   and ki67 staining of pancreatic sections in α-NIK-OE mice and control
   littermates (n = 5-6/group). D, Serum amylase levels (n = 7/group). E,
   Pancreatic trypsin activity (n = 8/group). F, Representative H&E
   staining of pancreatic sections in α-NIK-OE mice and control
   littermates. Note the increased edema (black stars) in α-NIK-OE mice.
   G, F4/80 and CCL2 were co-immunostained in the pancreatic sections of
   α-NIK-OE mice and control littermates (n = 5-6/group). H, Pancreas mRNA
   abundance was measured by RT-qPCR and normalized to 36B4 levels (n =
   9-12/group). I, Serum chemokine (CCL2 and CCL5) levels were measured by
   ELISA (n = 8-10/group). *, p < 0.05. **, p < 0.01.

Discussion

   In this study, we investigated the role of NIK in the regulation of
   islet α cells and showed that islet α-cell-specific overexpression of
   NIK induces islet α-cell dysfunction. Importantly, islet
   α-cell-specific overexpression of NIK results in islet β-cell death and
   pancreatitis, which is most likely due to paracrine secretion of
   cytokines and chemokines from the affected islet α cells. Ultimately,
   this islet α-cell dysfunction leads to hypoglycemia, growth
   retardation, and postnatal death in these mice. These observations
   indicate that islet α-cell inflammation induced by NIK is directly
   linked to pancreatitis.

   In support of these conclusions, we first showed that the
   LTβR/noncanonical NF-κB signaling pathway can be activated in
   pancreatic α cells. Stimulation of LTβR by an anti-LTβR antibody or
   LIGHT is able to activate the noncanonical NF-κB signaling pathway in
   islet α cells. As NIK is the key activator of the noncanonical NF-κB
   signaling pathway [166]^4, overexpression of NIK in islet α cells
   results in impaired glucagon secretion, both in vitro and in vivo,
   which is at least partially due to decreased glucagon expression. Both
   NIK and p52 could dose-dependently inhibit glucagon promoter Luciferase
   activity, indicating that NIK/p52 directly suppresses glucagon gene
   transcription. Overexpression of NIK also causes islet α-cell death and
   reduced cell proliferation, leading to a decreased islet α-cell mass,
   further contributing to decreased serum glucagon levels in α-NIK-OE
   mice. Insufficient secretion of glucagon, thus, contributes to the
   observed hypoglycemia in these mice.

   In addition to islet α-cell phenotypes, we also observed islet β-cell
   phenotypes in α-NIK-OE mice. Serum insulin levels and the islet β-cell
   mass were significantly decreased in α-NIK-OE mice due to increased
   islet β-cell death and decreased islet β-cell proliferation. This
   crosstalk between islet α and β cells is further supported by the
   results of our conditioned medium experiments. NIK overexpression in
   islet α cell induces inflammatory cytokine and chemokine expression,
   which are then secreted from islet α cells and act in a paracrine
   manner on islet β cells, causing islet β-cell death and reduced β-cell
   proliferation. Finally, these results in lower serum insulin levels,
   which may contribute to the growth retardation observed in α-NIK-OE
   mice.

   Because glucagon deficiency and islet-cell dysfunction do not cause
   postnatal death in mice [167]^18^, [168]^19, we further assessed the
   function of the pancreas, liver, and intestine of α-NIK-OE mice. It was
   observed that only pancreas develops severe pathological phenotypes
   (pancreatitis). This is the first evidence that islet α-cell
   inflammation can cause pancreatitis. Consistent with the association
   between islet β-cell stress in type 2 diabetes mellitus and the
   development of pancreatitis [169]^3, islet α cells overexpressing NIK
   releases inflammatory cytokines and chemokines, which may further
   damage acinar cells and recruit macrophage into the pancreas, inducing
   inflammation and pancreatitis. Subsequently, hypoglycemia and
   pancreatitis ultimately result in growth retardation and postnatal
   death in α-NIK-OE mice.

   In conclusion, we have uncovered a role for NIK in the regulation of
   islet α cells and found that islet α-cell-specific overexpression of
   NIK both induces islet α cell dysfunction and causes islet β cell death
   resulting in pancreatitis, which is most likely due to paracrine
   secretion of cytokines and chemokines from islet α cells, resulting in
   hypoglycemia, growth retardation, and postnatal death in these mice.
   These data suggest for the first time a relationship between islet
   α-cell inflammation and the development of pancreatitis.

Acknowledgments