Role of O-GlcNAcylation and endoplasmic reticulum stress on obesity and insulin resistance

Animals

Genetically obese mice C57BL/6 Lep^ob/Lep^ob (ob/ob) and as a control wild-type littermates (purchased from Charles River) used to determine O-GlcNAc modification status in the context of insulin resistance and ER stress. Mice were fed standard chow (24% protein, 18% fat, 58% carbohydrate) and had free (ad libitum) access to food. The animals were kept on a 12 h/12 h light/dark cycle.

Four weeks old male mice were used for experiments. Body weight of the mice recorded weekly up to 12 weeks. Then to evaluate peripheral insulin resistance in the mice the Intraperitoneal Glucose Tolerance Test (IPGTT) is performed at the 12th week as previously described (8). Briefly, animals fasted for 16 h and blood was collected from the tail vein to determine fasting glucose levels. Ob/ob (n=5) and wild-type (Lean) (n=5) mice then received 2 mg/g body weight of a sterile glucose solution by intraperitoneal injection. Blood samples were collected at 0, 15, 30, 60, 120, and 180 min following the injection and circulating glucose was measured using the Lifescan (Mountain View, CA, USA) One Touch monitor. According to the blood glucose values against time plotting the area under the curve calculated and the results interpreted. Blood glucose value for every time point normalized to the value of time zero. Plasma samples prepared from collected blood at the beginning of IPGTT (time zero) and plasma insulin levels measured by ELISA (Elabscience, mouse insulin ELISA kit). To further evaluation of peripheral insulin resistance level of the mice HOMA-IR calculated according to the formula: fasting insulin (μU/L)×fasting glucose (nmol/L)/22.5. According to the test results, mice found to be insulin resistant killed by cervical dislocation, visceral and epididymal adipose tissues collected. To understand the effect of external stimulation of O-GlcNAc on insulin signaling pathway and UPR wild type mice assigned to either saline or 5 mg/kg/day glucosamine for 5 days infusion groups. Glucosamine hydrochloride stock solution prepared as previously described briefly, 200 mg/mL in PBS [23]. Male mice (n=4) injected intraperitoneally every 24 h for 5 days with glucosamine male mice in the control group (n=4) injected at the same time and the dose of sterile PBS. At the end of the injections, IPGTT performed to determine the effect of glucosamine injection to peripheral insulin resistance. Adipose tissues collected from the mice after cervical dislocation and stored at −80°C until further analyses.

Preparation of protein extracts and Western blotting

Collected adipose tissues lysed in modified RIPA lysis buffer (200 mM NaCl, 50 mM Tris-HCl pH8, 0.05% SDS, 2 mM EDTA, 1% NP40, 150 mM sodium chloride, 1% deoxycholate including protease and phosphatase inhibitors, 1.40 mM glucosamine, 0.2 μM PUGNAc). Total protein lysates quantified using the bicinchoninic acid (BCA) method (Thermo Scientific, Pierce) and normalized according to the manufacturer’s instructions. Western blotting performed according to standard procedure briefly, total protein lysates loaded on a 10% SDS-polyacrylamide gel and transferred to polyvinylidene fluoride (PVDF) and then membranes blocked in 3% bovine serum albumin phosphate buffered saline 0.1% Tween-20 solution (BSA-PBST). The following antibodies were used: BiP (GRP78) Cell Signaling #3183S, XBP1 Santa Cruz #sc-7160, Phospho-PERK (Thr⁹⁸⁰) Cell Signaling #3179S, Total PERK Cell Signaling #5683P, pelF2 α (Ser⁵¹) Cell Signaling #3597S, Total elF2 α Cell Signaling #5324P, ATF6 Elabscience, p-IRS-1 (Ser³⁰⁷) Cell Signaling #2381, p-IRS-1 (Ser⁶¹²) Cell Signaling #3203, p-IRS-1 (Ser³¹⁸) Cell Signaling #5610, Total IRS-1 Cell Signaling #3407, Total IRS-2 Cell Signaling #4502, p-AKT (Ser⁴⁷³) Cell Signaling #4060P, p-AKT (Thr³⁰⁸) Cell Signaling #13068P, Total AKT Cell Signaling #4685S, IR β Cell Signaling #3025S, CTD 110.6 Cell Signaling #9875S, RL2 Abcam #92858, DM17 (OGT) Sigma #O6264, 345 (OGA) Sigma #SAB4200267, GFAT (Elabscience), UAP1(Elabscience), GlcNE (Elabscience), β tubulin Cell Signaling #5666P. Secondary antibodies included: horseradish peroxidase-conjugated anti-mouse IgM, anti-mouse IgG (Cell Signaling #7076P2, and anti-rabbit IgG (Bio-rad 1706515). Immunoblots were scanned and quantified using Li-Cor Odyssey^® Fc Imaging System.

Statistical analysis

Data were analyzed using two-tailed Student’s t-test and one-way ANOVA using GraphPad Prism 7 software. Values with p<0.05 were considered significant.

Status of O-GlcNAc modification and hexosamine biosynthetic pathway (HBP) in ob/ob mice

Previous researchers have been shown that insulin resistance correlated with increased O-GlcNAc modification levels in various tissues [24], [25], [26], [27], [28], [29]. Genetically obese, hyperphagic ob/ob mice are a widely used model to study obesity and obesity-associated complications; however, it is not considered to examine adipose tissues of this model for O-GlcNAc modification status. To better understanding of the relationship between O-GlcNAc modification and insulin resistance, O-GlcNAc levels investigated in visceral and epididymal adipose tissues of obese and insulin resistant ob/ob mice. First of all, the Intraperitoneal Glucose Tolerance Test (IPGTT) and HOMA-IR calculations were performed to ensure that the mice were insulin resistant. IPGTT performed at the 12th week and revealed a delayed glucose clearance in ob/ob mice (Supplementary Figure 1A). Furthermore, fasting blood glucose levels, serum insulin levels and HOMA-IR values were significantly higher in ob/ob mice (Supplementary Figure 1B). In conclusion, ob/ob mice clearly different than wild-type littermates and considered as a insulin resistant.

To further assess tissue-specific insulin resistance, we examined the expression of total AKT and pAKT Ser⁴⁷³ in epididymal and visceral adipose tissues of ob/ob and wild-type mice by western blot. AKT phosphorylation is an important step in the insulin receptor-signaling cascade. The pAKT/Total AKT ratio is an important technique to determine the severity of insulin resistance. The lower it is the ratio, the stronger the insulin resistance. Diminished pAKT Ser⁴⁷³ and Thr³⁰⁸ levels found in visceral and epididymal adipose tissues of ob/ob mice, individually or in pooled samples (Figure 1). In conclusion, evaluation of tissue-specific level of insulin resistance revealed that visceral and epididymal adipose tissues of ob/ob mice are also insulin resistant.

Figure 1:

Glucose homeostasis in ob/ob and lean mice.

Lean control and ob/ob mice used to examine essential markers of insulin resistance in visceral and epididymal adipose tissues in pooled samples. Insulin receptor signaling markers including phosphorylated Ser⁴⁷³, Thr³⁰⁸, and total AKT; phosphorylated IRS1 Ser³⁰⁷, Ser³¹⁸, Ser⁶¹², and total IRS1, IRS2, Pro-insulin receptor, Insulin receptor β examined by immunoblotting. Quantification of IRS1 Ser³⁰⁷, Ser³¹⁸, and Ser⁶¹² phosphorylation normalized to total IRS1; AKT Ser⁴⁷³ and Thr³⁰⁸ normalized to total AKT protein levels. IRS1, IRS2, Pro-IR, IR-β normalization to tubulin. Data are shown as means±SEM. Statistical significance in two-tailed Student’s t-test indicated by *p≤0.05, **p≤0.005, and ***p≤0.0005.

Insulin receptor signaling is defective in adipose tissues of obese mice

As a well-described fact, obese mice develop insulin resistance [7], [8]. First, we confirmed that ob/ob mice have a defective insulin signaling pathway by evaluation of cascade members expression in the protein level by western blot. Recent studies showed that various pathophysiological conditions such as inflammation activate stress kinases; JNKs and inhibitor of nuclear factor kappa-B kinase subunit beta (IKKβ) leads to serine phosphorylation of insulin signaling cascade – instead of tyrosine – results in the development of insulin resistance. As a result of insulin resistance, serine phosphorylation of the insulin receptor substrates-1 (IRS-1) Ser³⁰⁷, Ser³¹⁸, Ser⁶¹² increased while total IRS1 and IRS2 levels slightly decreased in visceral and epididymal adipose tissues of ob/ob mice. The pro-insulin receptor β levels increased while insulin receptor β levels remained same in the visceral and epididymal adipose tissues of ob/ob mice (Figure 1). Thus, ob/ob mice developed insulin resistance and insulin receptor signaling profound defective.

Unfolded protein response increased in insulin resistance

ER stress is a crucial contributor in obesity-related insulin resistance [8]. Therefore, we examined in visceral and epididymal adipose tissues of insulin-resistant ob/ob mice whether developed ER stress. GRP78 is a major ER chaperone protein and increased GRP78 levels observed in visceral and epididymal adipose tissues of ob/ob mice. XBP1s and ATF6 are different branches of the UPR and critical transcription factors that regulate ER stress-related genes expression and their levels elevated in visceral and epididymal adipose tissue of ob/ob mice. Moreover, translation control branch of the UPR, PERK Thr⁹⁸⁰, and eIF2α Ser⁵¹ phosphorylations are elevated in visceral and epididymal adipose tissues of ob/ob mice. According to the ATF6 and XBP1 levels in epididymal adipose tissue is more stressed than visceral adipose tissue in ob/ob mice. Together, these findings revealed that in adipose tissues of ob/ob mice developed insulin resistance-related ER stress (Figure 2).

Figure 2:

Endoplasmic reticulum stress indicators elevated in obesity.

Genetic (ob/ob) mouse model of obesity used to examine endoplasmic reticulum stress in adipose tissue compared with age and sex-matched lean controls. Endoplasmic reticulum stress markers including PERK phosphorylation (Thr⁹⁸⁰) and eIF2α phosphorylation (Ser⁵¹), total PERK, total eIF2α, GRP78, XBP1, ATF6 levels were examined by immunoblotting in the visceral and epididymal adipose tissues. Quantification of PERK Thr⁹⁸⁰ and eIF2α Ser⁵¹, normalized to total protein and GRP78, XBP1, ATF6, eIF2α, PERK normalized to tubulin.

O-GlcNAcylation and hexosamine biosynthesis pathway affected by insulin resistance

It is known that phosphorylation shift occurs in proteins with insulin resistance. Thus, we aimed to investigate whether the status of O-GlcNAcylated proteins affected by insulin resistance in obesity. Analyses with RL2 and CTD110.6 antibodies specific to O-GlcNAc groups of proteins revealed that O-GlcNAc levels strikingly decreased in visceral and epididymal adipose tissues of the insulin-resistant ob/ob mice. Regarding that OGT, the enzyme responsible for attaching O-GlcNAc groups to proteins diminished comparatively. Moreover, hexosamine biosynthesis pathway components GFAT, UAP1, and OGA levels decreased, however, GlcnE levels increased in visceral and epididymal adipose tissue of ob/ob mice. In conclusion, the control mechanism of O-GlcNAcylation influenced by the malfunction of glucose metabolism (Figure 3). Taken together, visceral and epididymal adipose tissues of ob/ob mice exhibited insulin resistance and ER stress, although, interestingly, O-GlcNAc modification levels diminished and also HBP elements expression levels comparable in ob/ob mice.

Figure 3:

Alteration of the hexosamine biosynthetic pathway (HBP) in obesity.

Lean control and ob/ob mice were used to examine HBP markers in visceral and epididymal adipose tissues compared with age and sex-matched lean controls. O-GlcNAc modification visualized by two different antibodies (RL2 and CTD110.6) which recognize specifically O-GlcNAc group on the proteins. Hexosamine biosynthetic pathway members GFAT, OGA, OGT, GLCNE, UAP1 were examined by immunoblotting. Quantification of HBP, normalization to every protein to tubulin.

Glucosamine induced O-GlcNAc modification, insulin signaling, ER stress and HBP in WT mice

Glucosamine-induced O-GlcNAc modification leads to insulin resistance and ER stress

Glucosamine (GlcN), can be metabolized via the hexosamine biosynthesis pathway through bypassing the rate-limiting enzyme of the HBP, glutamine:fructose-6-phosphate amidotransferase (GFAT) therefore it is a potent stimulator of HBP pathway activity [30]. So, we investigated how glucosamine challenge can affect insulin signaling and UPR in wildtype mice. C57BL/6 mice were injected intraperitoneally with 5 mg/kg/day glucosamine for 5 days. First, we performed IPGTT to understand whether the role of glucosamine challenge on whole-body glucose tolerance. Interpretation of IPGTT results indicated that glucosamine challenged wild-type mice exhibited insulin resistance (Supplementary Figure 2). After that, we examined the O-GlcNAcylation levels in adipose tissue lysates by the RL2 antibody, as expected O-GlcNAcylation increased in wild-type mice either individual or pooled samples (Figure 4). Next, we confirmed that glucosamine effectively stimulates the HBP through enhanced expressions of OGA, OGT, GlcNE, UAP1 in adipose tissues of wild-type mice. However; the rate-limiting enzyme of the HBP, GFAT diminished in glucosamine challenged wild-type mice presumably due to the negative feedback inhibition (Figure 4). In conclusiıon, glucosamine-induced stimulation of O-GlcNAc modification of proteins via modulation of HBP members accomplished in wild-type mice. Subsequently, the observation of peripheral glucose intolerance by IPGTT, we investigated the expression of the insulin signaling pathway members to assess further how insulin signaling regulated in glucosamine-challenged wild type mice. We observed that glucosamine-challenged mice displayed enhanced phosphorylation of serine residues Ser³⁰⁷, Ser³¹⁸, and Ser⁶¹² in IRS1 furthermore, total IRS1, and IRS2 levels were also comparable. Moreover, p-AKT Ser⁴⁷³ and Thr³⁰⁸ decreased, pro-insulin receptor β levels increased however, insulin receptor β levels remained the same in glucosamine challenged wild-type mice (Figure 5). In conclusion, the external stimulation of O-GlcNAc through glucosamine interrupted insulin signaling and led to insulin resistance in wild type mice.

Figure 4:

In vivo Glucosamine infusion increases O-GlcNAc modification.

Adipose tissue of GlcN injected C57/BL6 mice were used to examine O-GlcNAc modification compared with PBS injected, age and sex-matched controls. O-GlcNAc modification visualized by two different antibodies (RL2 and CTD110.6) which recognize specifically O-GlcNAc group on the proteins. In vivo Glucosamine infusion increases O-GlcNAc modification and alters the hexosamine biosynthetic pathway (HBP) members. GFAT, OGA, OGT, GLCNE, UAP1 were examined by immunoblotting. Quantification of O-GlcNAc and HBP, normalization to every protein to Tubulin.

Figure 5:

In vivo GlcN infusion increases insulin resistance.

GlcN injected C57/BL6 mice were used to examine markers of insulin receptor signaling in adipose tissues compared with age and sex-matched lean controls. Insulin receptor signaling markers including phosphorylated AKT Ser⁴⁷³, Thr³⁰⁸, phosphorylated IRS1 Ser³⁰⁷, Ser³¹⁸ and Ser⁶¹², total IRS1, IRS2, proinsulin receptor and insulin receptor β levels examined by immunoblotting. Quantification of IRS1 Ser³⁰⁷, Ser³¹⁸, and Ser⁶¹² phosphorylations normalized to total IRS1, AKT Ser⁴⁷³, Thr³⁰⁸ normalized to total AKT, IRS1, IRS2, Pro-IR, IR-β normalization to tubulin.

Glucosamine induces ER stress

To understand how external O-GlcNAc stimulation effects ER stress we investigated UPR markers in adipose tissues of glucosamine-challenged wild type mice. We found that enhanced expression of GRP78, ATF6, and XBP1 regarding that phosphorylation of PERK Thr⁹⁸⁰ and eIF2α Ser⁵¹ in adipose tissue of glucosamine-challenged mice were also comparable (Figure 6). So, we can conclude that external O-GlcNAc stimulation not only leads to insulin resistance but also leads to ER stress in adipose tissue of wild type mice.

Figure 6:

Endoplasmic reticulum stress indicators increased with in vivo GlcN infusion.

C57/BL6 mice were used to examine endoplasmic reticulum stress markers of in adipose tissue compared with age and sex-matched lean controls. PERK phosphorylation (Thr⁹⁸⁰) and eIF2α phosphorylation (Ser⁵¹), total PERK, total eIF2α, GRP78, XBP1, ATF6 levels were examined by immunoblotting in the visceral and epididymal adipose tissues. Quantification of PERK Thr⁹⁸⁰ and eIF2α Ser⁵¹, normalized to total protein and GRP78, XBP1, ATF6, eIF2a, PERK normalized to Tubulin.

O-GlcNAc modification is essential in the development of insulin resistance

To further examine the interaction of the insulin receptor signaling and O-GlcNAc modification, we immunoprecipitated the IRS1 from adipose tissues of both ob/ob and glucosamine-challenged wild-type mice. After that, immunoprecipitated samples immunoblotted with anti-O-GlcNAc antibody CTD110.6, p-Tyr, and p-Ser/Thr antibodies, respectively. Tyrosine phosphorylation of IRS1 diminished; in contrast, serine-threonine phosphorylations of IRS1 increased in visceral and epididymal adipose tissues of ob/ob mice. In parallel with phosphorylation shift, O-GlcNAcylation of IRS-1 also increased in both visceral and epidydimal adipose tissues of ob/ob mice (Figure 7A). Similarly, tyrosine phosphorylation of IRS1 decreased furthermore, serine-threonine phosphorylations of IRS1 increased in the adipose tissue of glucosamine-challenged wild-type mice. In parallel with phosphorylation shift, O-GlcNAcylation of IRS-1 also increased in adipose tissues of glucosamine-challenged wild-type mice. Taking together, the IRS1 modified with O-GlcNAc in adipose tissue of ob/ob and glucosamine-challenged mice, and comparable with the serine phosphorylation suggesting that O-GlcNAcylated IRS-1 crucial factor for the development of insulin resistance.

Figure 7:

O-GlcNAcylation, tyrosine phosphorylation, serine phosphorylation of IRS1 in adipose tissue.

(A) IRS1 tyrosine phosphorylation, serine phosphorylation, and O-GlcNAcylation and their total protein levels of ob/ob mice were examined with immunoprecipitation (IP) followed by immunoblotting (IB) in adipose tissues. Quantification of IRS1 tyrosine and serine phosphorylations and O-GlcNAcylation with normalization to IRS1 protein levels. (B) In vivo GlcN infusion and O-GlcNAcylation, tyrosine phosphorylation, serine phosphorylation of IRS1 in adipose tissue of C57/BL6 mice. IRS1 tyrosine phosphorylation, serine phosphorylation, and O-GlcNAcylation and their total protein levels were examined with immunoprecipitation (IP) followed by immunoblotting (IB). Quantification of IRS1 tyrosine and serine phosphorylations and O-GlcNAcylation with normalization to IRS1 protein levels.