An important mechanism of pathogenesis of type 2 diabetes in obese individual is the elevation of free plasma fatty acids (FFA) and generation of reactive oxygen species (ROS). The objective of the current study is to determine whether the IKK/NF-kB inflammatory pathway also plays a role in liptoxicity. The IKK inhibitor salicylate was used in this study. Study was conducted in vivo using hyperglycemic clamp, ex-vivo with isolated islets, and in vitro.
Female Wistar Rats were infused with either a)oleate, b)olive oil+heparin, c)oleate+salicylate, d)olive oil + salicylate, e)salicylate, or f)saline for 48hrs. A two-step hyperglycemic clamp was followed or the islets were isolated. Islets were assayed for insulin and c-peptide.
Oleate and olive oil reduce GINF during the two-step hyperglycemic clamp. Oleate impaired insulin secretion when c-peptide was measured. This was prevented by the co-infusion of salicylate. Olive oil decreased insulin sensitivity when the sensitivity index was calculated. The effect of olive oil on insulin sensitivity is prevented by the co-infusion of salicylate. Ex-vivo and in vitro data showed similar result. The data here presented demonstrates the role of IKK/NF-kB in FFA induced liptoxicity and its prevention by salicylate.
Diabetes mellitus 2 is characterized by a combination of glucotoxicity and lipotoxicity. Evidences supporting glucotoxicity mediated decrease in β-cell function are well documented. However the mechanisms of free fatty acid (FFA) induced liptoxicity is not fully elucidated. FFA is needed acutely for glucose induced insulin secretion (GSIS). However, when chronically exposed to high levels of FFA, β-cell function decreases due to increased insulin resistance or impaired insulin secretion. In vitro results have shown that increased levels of FFA impair GSIS, in vivo results are more controversial with different groups presenting conflict results. Despite this mounting evidence suggest increased levels of FFA in obesity leads to insulin resistance. In vivo a decrease in insulin resistance should be accompanied by a corresponding increase in insulin secretion; this is not seen in individuals pre-disposed to type-2 diabetes, suggesting that other pathway may also be involved. One theory proposes that the release of adipose tissue mass is responsible for intracellular alternations that reduce insulin signaling. Adipose tissues are known to release cytokines such as TNF-α, which causes insulin resistance by serine phosphorylation of the insulin receptor and its substrate (IRS-1). Recently it has been shown that another pathway that elevated FFA may cause β-cell dysfunction is through the generation of reactive oxygen species (ROS).
Previous studies have demonstrated that islets cells are especially susceptible to oxidative stress due to inherent low anti-oxidant defense. Proposed pathways of liptoxicity such as activation of JNK, PKC, or IKK/NF-kB pathway are often either upstream or downstream of oxidative stress. Studies conducted in this lab have demonstrated that antioxidants such as taurine, N-Acetylcysteine (NAC), and TEMPOL are able to prevent β-cell dysfunction in rats with elevated FFA levels. Cytosolic superoxide is found to be increased in oleate infused rat while mitochondrial superoxide was not increased.
The antioxidants NAC and taurine does not decrease superoxide suggesting other ROS also play important role in lipotoxicity. These other reactive oxygen species can be detected by H2DCF-DA. The ability of reactive oxygen species to activate transcription factor NF-kB and the inflammatory pathway are well established. The next logical step would be to investigate whether FFA also induce activation of NF-kB by generating ROS.
The transcription factor NF-kB is activated downstream of ROS generated by elevated FFA. NF-kB is responsible for the transcription of various cytokines and chemokines that mediates inflammatory response. However, its close association with ROS would suggest that NF-kB also plays a role in lipotoxicity. Nuclear localization of NF-kB is dependent upon activation by IKKβ. Yuan and colleagues were able to demonstrate that by using salicylate (a weak inhibitor of IKKβ), diet induced hyperglycemia and dyslipidemia could be reversed. The paper demonstrated that by inhibiting the IKK/NF-kB pathway insulin resistance could be reversed. Studies have also shown the ability of salicylate to prevent hepatic and peripheral insulin resistance in rat models. Liptoxicity is a combination of insulin resistance and dysfunction in insulin secretion. A decrease in insulin sensitivity should be compensated by an increase in insulin secretion, but this is not seen in models of lipotoxicity. Hence it is also possible that the NF-kB pathway is involved in insulin resistance and decreased insulin secretion at the β-cell level. The aim of the study presented here is to determine whether FFA induced oxidative stress also activate NF-kB within islets to cause defect in insulin sensitivity and insulin secretion. An in vivo, ex-vivo, and in vitro models are presented here. The IKKβ inhibitor salicylate will be used to investigate the possibility of NF-kB activation within this pathway. This study will help to further elucidate mechanisms of liptoxicity and present possible therapeutic options.
Female Wistar rats were cannulated 2-3 days prior infusion. During surgery rats were anesthetized with isofluorane. Catheters were injected into right jugular vein for infusion and left carotid artery for sampling. All procedures are approved by the University of Toronto Animal Care facility.
Intravenous Infusion: Rats were infused for 48 hours with 1)saline; 2)oleate; 3)olive oil; 4)oleate and salicylate; 5)olive oil and salicylate; 6)salicylate, followed by a two-step hyperglycemic clamp. Same infusion protocol was followed for ex-vivo study. Olive oil was prepared from a triglyceride mixture containing 20% olive oil infused at 5µl/min and heparin at 50u/ml. Olive oil with salicylate was infused at a rate of 5µl/min. Oleate was infused at 5µl/min. Oleate with salicylate was infused at 5µl/min. Salicylate was infused at 10µl/min and saline was used as control with an infusion rate of 5µl/min.
Hyperglycemic clamps: A two-step hyperglycemic clamp (13mmol/l and 22mmol/l) was used to determine plasma insulin and c-peptide content. An infusion of 37.5% glucose was started at time zero minute. Plasma glucose was maintained at 13mmol/l by adjusting the rate of glucose infusion every 10 minutes. After two hours the plasma glucose level was raised to 22mmol/l and maintained until the end of the experiment (4 hours). Plasma glucose levels were analyzed using Beckman Glucose Analyzer. Radioimmunoassay for c-peptide and insulin was used to determine plasma concentration.
Ex vivo and in vitro studies: After 48hr infusion pancreas, the rat were cut open at the abdominal section and infused with collagenase in the pancreas. Pancreas was then placed in a 50ml test tube and incubated at 37⁰C. The contents were subsequently centrifuged and islets were isolated by Histoplaque density gradient. After pre-incubation in KRBB, islets were incubated at 2.8mM, 6.5mM, 13mM, and 22mM of glucose. For in vitro study cells were cultured in RPMI1640. Oleate was added at 0.4mM in 0.5% BSA.
In-vivo clamp studies: After 48hr infusion a two-step hyperglycemic clamp was performed to evaluate insulin secretion. During the first step the glucose level was elevated to 13mmol/l the upper physiological concentration in rats. In the second phase glucose level was elevated to 22mmol/l to achieve maximal stimulatory response. After 48hr infusion plasma FFA were elevated 2 fold with either oleate or olive oil. Basal plasma glucose levels were similar in all rats prior to the first infusion period. The glucose infusion rate (GINF) need to maintain the target the glucose rate was lower in oleate and olive oil infusion compared to the control (Figure 1a, 1b). This is consistent with decreased insulin secretion caused by oleate and decreased insulin sensitivity caused by olive oil. Infusion of oleate with salicylate or olive oil with salicylate prevented the effect of lowered GNIF cause by either fat.
Figure 1: Effect of oleate or olive oil on glucose infusion rate (GINF) during a two step hyperglycemic clamp. A) Oleate infusion resulted in lower GNIF and is restored by co-infusion of salicylate. B) Infusion of olive oil resulted in lower GNIF and is restored by co-infusion of salicylate. (p<0.01)
Basal Insulin and C-peptide levels were similar in all groups. As expected plasma insulin rose in response to increasing glucose levels (Figure 2a, 2b). Plasma c-peptide, derived from cleavage or pro-insulin, also rose indicating that the rise in insulin content was due to increased secretion (Figure 3a, 3b). Plasma insulin was lower in oleate treated rats compared with saline treated rats. The effect of reduced insulin secretion caused by oleate can be prevented by the co-infusion of salicylate with oleate as shown by the oleate and salicylate treated rats. The co-infusion of salicylate completely restored insulin content at 13mmol/l of glucose and 22mmol/l of glucose. In contrast olive oil did not decrease plasma insulin content or c-peptide level when compared to saline treated rats. Co-infusion of salicylate with olive oil also had no effect when compared to the control.
Figure 2: Plasma insulin and c-peptide secretion during the hyperglycemic clamp in 12 week old normal female Wistar rats treated for 48h with an i.v. infusion of: A) 1) Saline alone (SAL n=12), 2) Oleate alone (OLE 1.3µEq·min-1 n=10), 3) Oleate + Salicylate (OLE+SLY, oleate-1.3µEq·min-1, SLY-0.117mg·kg-1min-1, n=8), 4) Salicylate alone, (SLY, n=9); B) 1) Saline alone (SAL n=12), 2) Olive oil alone (OLO 5µl·min-1 n=7), 3) Olive oil + Salicylate (OLO+SLY, olive oil-5µl·min-1, SLY-0.117mg·kg-1min-1, n=11), 4) Salicylate alone, (SLY, n=9); Data are means ± SE. C) 1) Saline alone (SAL n=12), 2) Oleate alone (OLE 1.3µEq·min-1 n=10), 3) Oleate + Salicylate (OLE+SLY, oleate-1.3µEq·min-1, SLY-0.117mg·kg-1min-1, n=8), 4) Salicylate alone, (SLY, n=9); D) 1) Saline alone (SAL n=12), 2) Olive oil alone (OLO 5µl·min-1 n=7), 3) Olive oil + Salicylate (OLO+SLY, olive oil-5µl·min-1, SLY-0.117mg·kg-1min-1, n=11), 4) Salicylate alone, (SLY, n=9); Data are means ± SE. (p<0.01)
Ex-vivo studies in Islets: Rats were infused for 48hrs with respective treatments at the same rate as in the in-vivo study. After 48hrs islets were isolated and incubated in 2.8mmol/l (non-stimulatory), 6.5mmol/l (basal glucose level in rats), 13mmol/l and 22mmol/l (hyperglycemic clamp) of glucose (Figure 3a, 3b). GSIS in olive oil treated rats (0.996+0.132 pmol/islet/h at 22mM of glucose versus control 1.542+0.159 pmol/islet/h at 22mM) was impaired compared to saline treated rats. GSIS in oleate treated rats were also impaired when compared to control (1.033+0.097 pmol/islet/h at 22mM versus control: 1.542+0.159 pmol/islet/h at 22mM). Co-infusion of salicylate prevented the negative effects of olive oil (olive oil + salicylate: 1.754+0.367 pmol/islet/h at mM) and oleate (oleate + salicylate: 1.737+ 0.311 pmol/islet/h at 22mM) on GSIS. Salicylate alone had no effect (1.524+0.218 pmol/islet/h at 22mM).
Figure 3. Insulin secretory response to glucose of freshly isolated islets of 12 week old normal female Wistar rats treated for 48h with: A) 1) Saline alone (SAL n=16), 2) Oleate alone (OLE 1.3 µEq·min-1 n=14), 3) Oleate + Salicylate (OLE+SLY, oleate-1.3µ Eq·min-1, SLY-0.117 mg·kg-1min-1, n=8), 4) Salicylate alone, (SLY, n=10); B) 1) Saline alone (SAL n=16), 2) Olive oil alone (OLO 5µl·min-1 n=12), 3) Olive oil + Salicylate (OLO+SLY, olive oil-5µl·min-1, SLY-0.117mg·kg-1min-1, n=6), 4) Salicylate alone, (SLY, n=10); (p<0.01)
In-vitro studies: In-vitro study was carried out to investigate if 48hr infusion of oleate or co-infusion with salicylate had the same effect as in in-vivo or ex-vivo study. This would also determine whether the oleate induced impairment of GSIS is solely prevented by the addition of salicylate or whether other physiological factors also play an important role in this pathway. In-vitro, similar to in-vivo and ex-vivo, oleate impaired GSIS when compared to control (oleate: 0.212+0.026 pmol/islet/h at 22mM versus control: 0.396+0.033 pmol/islet/h at 22mM) (Figure 4). Addition of salicylate in oleate cell culture prevented the effect of oleate and restored GSIS compared to control (oleate + salicylate: 0.391+0.032 pmol/islet/h at 22mM versus control: 0.396+0.033 pmol/islet/h at 22mM). Salicylate alone had no effect (0.416+0.065 pmol/islet/h at 22mM).
Sensitivity Index and Disposition Index: The sensitivity index and disposition index were calculated for the two-step hyperglycemic clamp. Sensitivity index is calculated GINF divided by insulin content. 48 infusion of oleate did not have any effect on the sensitivity index when compared with saline (Figure 5a). As the two step hyperglycemic clamp demonstrated, 48hr infusion of olive oil did not impair the absolute insulin secretion as measured by the c-peptide level. This is because olive oil infusion decreased insulin sensitivity when measured by the sensitivity index (Figure 5b). The infusion of salicylate with olive oil prevented the decrease in sensitivity index. Salicylate alone had no effect on insulin sensitivity. In vivo β-cell compensates for insulin resistance by increasing insulin secretion. The disposition index calculated as the sensitivity index multiplied by c-peptide level is conventionally used as an index of β-cell function. When disposition index was used to correct for β-cell sensitivity, both oleate infusion and olive oil infusion resulted in a decrease in DI compared with saline (Figure 6a, 6b). This is consistent with the reduced β-cell function caused by oleate or olive oil. The infusion of salicylate restored the disposition index. Salicylate alone had no effect on the disposition index.
Figure 5. Sensitivity Index (M/I = GINF/Insulin) during the hyperglycemic clamp in 12 week old normal female Wistar rats treated for 48h with an i.v. infusion of: A) 1) Saline alone (SAL n=12), 2) Oleate alone (OLE 1.3µEq·min-1 n=10), 3) Oleate + Salicylate (OLE+SLY, oleate-1.3µEq·min-1, SLY-0.117mg·kg-1min-1, n=8), 4) Salicylate alone, (SLY, n=9); B) 1) Saline alone (SAL n=12), 2) Olive oil alone (OLO 5µl·min-1 n=7), 3) Olive oil + Salicylate (OLO+SLY, olive oil-5µl·min-1, SLY-0.117mg·kg-1min-1, n=11), 4) Salicylate alone, (SLY, n=9); Data are means ± SE.
Figure 6. Disposition Index (DI = M/I x C-peptide, index of β-cell function) during the hyperglycemic clamp in 12 week old normal female Wistar rats treated for 48h with an i.v. infusion of: A) 1) Saline alone (SAL n=12), 2) Oleate alone (OLE 1.3µEq·min-1 n=10), 3) Oleate + Salicylate (OLE+SLY, oleate-1.3µEq·min-1, SLY-0.117mg·kg-1min-1, n=8), 4) Salicylate alone, (SLY, n=9); B) 1) Saline alone (SAL n=12), 2) Olive oil alone (OLO 5µl·min-1 n=7), 3) Olive oil + Salicylate (OLO+SLY, olive oil-5µl·min-1, SLY-0.117mg·kg-1min-1, n=11), 4) Salicylate alone, (SLY, n=9); Data are means ± SE.
This study examines the effect of prolonged exposure to FFA on the NF-kB activated inflammatory pathway. In vivo, infusion of oleate impaired insulin secretion as measured by c-peptide, while olive oil did not impair insulin secretion but impaired insulin sensitivity. This would suggest that different fat impair β-cell function through different mechanisms. However, the effect of oleate and olive oil were prevented by the co-infusion of salicylate.
Salicylate is a weak inhibitor of IKK. The infusion of salicylate with oleate prevented the reduction in insulin secretion caused by oleate. The co-infusion of salicylate with olive oil also prevented the reduction in insulin resistance. In this study a two-step hyperglycemic clamp was used to calculate insulin sensitivity, however the gold standard for determining insulin sensitivity is the hyperinsulinemic-euglycemic clamp. A hyperinsulinemic-euglycemic should also be performed to validate the effects of olive oil. Despite this, the study demonstrates that even though oleate and olive oil are different types of fat which negatively affect different aspects of β-cell functions, both are mediated by the IKK/NF-kB activated pathway. Ex-vivo and in vitro data show similar results. Salicylate was shown in both cases to prevent the effect of oleate and olive oil on insulin secretion and insulin sensitivity. In a proof of mechanism study, Dr. Fong of Prince Edward Island University was able to demonstrate oleate or olive oil increased the amount of IkB-α and phosphorylation of IKK. This is followed by the nuclear localization of NF-kB. However, the effect of IKK is prevented by salicylate and no NF-kB nuclear localization occurred. This is also consistent with the results of our findings. Further studies in mice also demonstrated that BMS, a much more specific IKK inhibitor than salicylate, was also able to prevent decrease in β-cell function caused by fat infusion (data now shown here). When disposition index was calculated, impairment of β-cell functions is found with both types of fat, but completely restored with salicylate. In theory a decrease in insulin sensitivity should be compensated by an increase in β-cell secretion. However this is not observed in the c-peptide measurements when olive oil was infused. (How do we explain this?) Previous findings in the lab demonstrated that FFA decreases GSIS by generating reactive oxygen species. Here we show that a possible downstream pathway is the activation of IKK/NF-kB. NF-kB is known to play an important role in activating cytokines and pro-inflammatory molecules. However, NF-kB is also able to act to promote transcription of non-inflammatory molecules. Further studies on the NF-kB activated genes would help elucidate the molecular mechanisms that attributes to decreased insulin sensitivity or secretion. Despite these findings, we should not quickly assume that salicylate or IKK inhibitors will work efficiently in a therapeutic setting. Mechanisms of FFA induced oxidative stress also function through other path ways such as JNK which are uninhibited by the addition of IKK inhibitors. Dr. Gary Lewis is currently researching the effect of salicylate in human models. This should not only provide therapeutic implications but also corroborate with the data presented. In conclusion this study demonstrates that one pathway of FFA induced lipotoxicity is through the activation of IKK/NF-kB pathway. This results in either a decrease in insulin secretion or insulin sensitivity which can be prevented by the IKK inhibitor salicylate.