Lipotoxicity of the Pancreatic β-Cell Is Associated With Glucose-Dependent Esterification of Fatty Acids Into Neutral Lipids

Reagents. Palmitic acid (sodium salt), palmitate methyl ester,octanoate, fatty-acid—free bovine serum albumin (BSA), triolein, and the GPO Trinder reagent (No. 339) were from Sigma (St. Louis, MO). 2-Bromopalmitate was from Aldrich (Milwaukee, WI). Reverse transcriptase—polymerase chain reaction (RT-PCR) primers and the Gold RT-PCR kit were from PerkinElmer (Foster City, CA).[U-¹⁴C]Palmitate, [1-¹⁴C]cholesteryl oleate, and[carboxyl-¹⁴C]triolein were from NEN (Boston, MA). Thesit was from Boehringer Mannheim (Mannheim, Germany).

Fatty-acid solutions. Stock solutions were prepared as follows:palmitic acid was dissolved in ethanol:H₂O (1:1, vol:vol) at 50°C at a final concentration of 150 mmol/l; 2-bromopalmitate and palmitate methyl ester were dissolved in ethanol at a final concentration of 300 mmol/l; and octanoate was dissolved in ethanol:H₂O (1:1,vol:vol) at a final concentration of 500 mmol/l. Aliquots of stock solutions were complexed with fatty-acid—free BSA (10% solution in H₂O)by stirring for 1 h at 37°C and then diluted in culture media. The final molar ratio of fatty acid:BSA was 5:1. The final ethanol concentration was≤0.33% (vol:vol). All control conditions included a solution of vehicle(ethanol:H₂O) mixed with fatty-acid—free BSA at the same concentration as the fatty-acid solution, unless indicated otherwise.

Animals. Six-week-old male Wistar rats were purchased from Harlan Sprague Dawley (Indianapolis, IN). Seven-week-old male Zucker diabetic fatty(ZDF/Gmi fa/fa) rats and Zucker lean control (ZLC/Gmi +/faor +/+) rats were purchased from Genetic Models (Indianapolis, IN). Animals were housed on a 12-h light/dark cycle with free access to water and standard laboratory food. All procedures using animals were approved by the Institutional Animal Care and Use Committee.

HIT-T15 cells and isolated rat islets. HIT-T15 cells (passage 69-78)were routinely cultured as described(11). For Northern blot,labeling, and TAG content experiments, cells were detached and subcultured for 48 h by plating 1 × 10⁶ cells per well in 6-well plates. Rat islets were isolated by collagenase digestion as described(3). After an overnight culture in RPMI 1640 containing 10% fetal bovine serum and 11.1 mmol/l glucose to ensure optimal recovery (12),batches of 100-200 islets were incubated in various experimental conditions as described in RESULTS.

Northern analysis of insulin mRNA. Cultured islets were transferred to 15-ml conical tubes, rinsed twice with phosphate-buffered saline (PBS) (137 mmol/l NaCl, 2.7 mmol/l KCl, 4.3 mmol/l NaH₂PO₄.7H₂O, pH 7.3), and resuspended in denaturing solution (4 mol/l guanidine thiocyanate, 25 mmol/l sodium citrate,pH 7, 0.5% sarcosyl, 0.1 mol/12-mercaptoethanol). HIT-T15 cells were rinsed with PBS and scraped with denaturing solution. RNA was extracted according to Chomczynski and Sacchi (13),and insulin mRNA was assessed by Northern blot as described(3). Membranes were stripped and rehybridized with a β-actin cDNA to control for variations in the amount of total RNA loaded on each lane.

Fluorescence-based RT-PCR of GPAT, DGAT, HSL, and insulin mRNAs.Comparative analysis of GPAT, DGAT, and HSL mRNA levels in isolated islets cultured with and without palmitate was performed using the Gold RT-PCR kit and an ABI Prism 7700 sequence detector equipped with a thermocycler (TaqMan technology), as described(14). β-Actin was used as an internal control, and insulin was used as a positive control for the effect of palmitate. Primer and probe sequences are given in Table 1.

Analysis of intracellular lipids by thin-layer chromatography.Fatty-acid solutions were prepared as described above with the addition of 5μmol/l (4.25 mCi) [U-¹⁴C]palmitate in all conditions. Islets or HIT-T15 cells were cultured in the various fatty-acid solutions, as described in RESULTS. At the end of the culture period, islets or cells were transferred to 12 × 75 mm prechilled glass tubes on ice, centrifuged at 1,200 rpm for 10 min at 4°C, washed with 2 ml ice-cold PBS, centrifuged as above,resuspended in 3 ml chloroform:methanol:HCl (200:100:1), and stored at 4°C overnight under N₂. After addition of 750 μl double-distilled H₂O, samples were vortexed and centrifuged for 10 min at 4°C. The upper (aqueous) phase was removed, and the lower (organic) phase was washed with 750 μl double-distilled H₂O and centrifuged as above. The upper phase was removed and the lower phase was dried under N₂. The lipid pellet was resuspended in 100 μl chloroform:methanol:HCl (200:100:1), and 40 μl was spotted in duplicate on silica gel thin-layer chromatography (TLC) plates (Whatman, Clifton, NJ). Lipids were resolved in petroleum ether:diethyl ether:glacial acetic acid(70:30:1). [1-¹⁴C]Cholesteryl oleate, [U-¹⁴C]palmitate,and [carboxyl-¹⁴C]triolein were used as standards. Results obtained with this system were initially verified using N-hexane:diethyl ether:methanol:glacial acetic acid (90:20:2:3). TLC plates were exposed to X-OMAT AR autoradiography films (Eastman Kodak, Rochester, NY). Individual bands were scraped, transferred to scintillation vials with 5 ml Ecoscint scintillation fluid (National Diagnostics, Atlanta, GA), and counted.

Intracellular TAG assay. We have experienced difficulties in trying to measure TAG content in islet extracts, as previously published(6,15,16,17,18,19). Therefore, we have modified this method by 1) extracting the lipids before the assay, 2) resuspending the extracted lipids in the detergent Thesit as recommended in other cells(20), and 3)calculating the sample TAG concentrations from a triolein standard curve, as opposed to a single glycerol value. The assay is based on the colorimetric determination of glycerol produced by hydrolysis of neutral lipids in the presence of lipoprotein lipase. Islets or cells were transferred to 12 ×75 mm prechilled glass tubes on ice, centrifuged at 1,200 rpm for 10 min at 4°C, washed with 500 μl ice-cold PBS, and centrifuged as above. The pellet was resuspended in 3 ml of chloroform:methanol (2:1) and stored overnight at 4°C under N₂. After the addition of 1.5 ml double-distilled H₂O, the tubes were vortexed and centrifuged at 1,200 rpm for 10 min at 4°C. The upper phase was discarded, and 750 μl double-distilled H₂O was added to the lower phase. The samples were vortexed and centrifuged at 1,200 rpm for 10 min at 4°C. The upper phase was removed, and the lower phase was evaporated under N₂. The samples were resuspended in 50 μl chloroform; 10 μl was quickly transferred to glass tubes in duplicate and air dried. The dry pellet was resuspended in 10 μl Thesit. The triolein standard curve (1-50 μg) was prepared using triolein (Sigma) diluted in chloroform:methanol (2:1). Samples and standards resuspended in Thesit were air dried; 50 μl H₂O was added; and the tubes were vortexed and incubated in a 37°C shaking water bath for 10 min as described(20). One milliliter of GPO Trinder reagent was added to the tubes, which were then gently mixed and incubated at 37°C for 5 min. The absorbance was read at 540 nm in a Beckman (Fullerton, CA) DU-64 spectrophotometer. To test the efficiency of TAG extraction, samples consisting of 5, 10, or 50 μg total lipid, each containing 0.9 μg of [carboxyl-¹⁴C]triolein, and the remainder unlabeled triolein underwent the extraction process. Aliquots were taken at each step during the extraction process and counted. The total loss of radioactivity during the extraction process was 3.0 ± 1.3% (n= 4).

Expression of data and statistics. Data are expressed as means± SE. Intergroup comparisons were performed by paired Student's t test or analysis of variance with Dunnett's t test for multiple comparisons, where appropriate. P < 0.05 was considered significant.

Expression of GPAT, DGAT, and HSL in isolated rat islets. We first sought to determine whether critical enzymes of neutral lipid metabolism are expressed in islets and whether their level of expression is modulated upon prolonged exposure to palmitate and glucose. GPAT is a rate-limiting enzyme for neutral lipid synthesis(21), and its expression is increased in islets from ZDF rats(6). DGAT is the enzyme responsible for the synthesis of TAG from diacylglycerols (DAGs), but its expression has never been documented in β-cells. HSL is responsible for the hydrolysis of TAG and is expressed and active in β-cells(22). Fluorescence-based RT-PCR demonstrated the expression of GPAT(Fig. 1A), DGAT(Fig. 1B), and HSL(Fig. 1C) in islets isolated from normal Wistar rats. Semiquantitative analysis of mRNA levels for each enzyme was then used to determine the effects of a 72-h culture in 2.8 or 16.7 mmol/l glucose with or without 0.5 mmol/l palmitate(Fig. 2). As expected, insulin mRNA levels were increased in the presence of 16.7 mmol/l glucose (P= 0.002, n = 8; Fig. 2A), and this effect was blocked by palmitate (P= 0.002, n = 8; Fig. 2A). However, neither GPAT(Fig. 2B), DGAT(Fig. 2C), nor HSL(Fig. 2D) mRNA was significantly affected by prolonged exposure to elevated glucose and/or palmitate. DGAT mRNA levels tended to be slightly lower in the presence of palmitate at 2.8 mmol/l glucose, but this difference was not statistically significant (NS, n = 7; Fig. 2C). The effect of the culture period itself was investigated by comparing insulin mRNA levels in freshly isolated islets to those in islets cultured for 72 h in either 2.8 or 16.7 mmol/l glucose. Insulin mRNA levels were lower in islets cultured in 2.8 mmol/l glucose than in freshly isolated islets (insulin:β-actin mRNA ratio 2.5 ± 0.9 vs. 5.7 ± 0.2, n = 3). Islets cultured for 3 days in 16.7 mmol/l glucose had insulin mRNA levels similar to those of freshly isolated islets (insulin/β-actin mRNA ratio 4.5 ± 1.5 vs. 5.7 ± 0.2, n = 3). Therefore, we cannot exclude the possibility that the lack of effect of palmitate on insulin mRNA levels at 2.8 mmol/l glucose is due to the fact that insulin mRNA is already downregulated at this glucose concentration and cannot be decreased further.

Effects of exogenous palmitate and glucose on incorporation of[¹⁴C]palmitate into neutral lipids. Intracellular neutral lipid content results from the balance between lipogenic (GPAT and DGAT) and lipolytic (HSL) enzyme activities. To determine whether neutral lipids accumulate upon prolonged exposure to elevated palmitate and glucose, isolated islets were cultured for 72 h in the presence of 5 μmol/l[U-¹⁴C]-labeled palmitate with and without addition of 0.5 mmol/l unlabeled palmitate, in 2.8 or 16.7 mmol/l glucose. Intracellular lipids were extracted and analyzed by TLC. In the absence of unlabeled palmitate,culturing islets with 16.7 mmol/l glucose increased the total number of counts incorporated from labeled palmitate into complex lipids (phospholipids [PLs] +DAG + TAG) 3.0 ± 0.9 fold compared with islets cultured in 2.8 mmol/l glucose (P < 0.05, n = 7; Fig. 3A). In islets cultured in 0.5 mmol/l unlabeled palmitate, the total number of counts incorporated into complex lipids was 2.4 ± 0.8-fold higher in the presence of 16.7 mmol/l glucose than in the presence of 2.8 mmol/l glucose(P < 0.05, n = 7; Fig. 3A). The ability of high glucose to increase incorporation of the counts into complex lipids was therefore similar in the absence or presence of unlabeled palmitate. The total number of counts incorporated into complex lipids was artificially lower in the presence of unlabeled palmitate because of the dilution of the specific activity of the tracer. The effect of palmitate on partitioning the counts is therefore expressed as the percentage of the counts incorporated into each lipid fraction (Fig. 3B). In the presence of 0.5 mmol/l unlabeled palmitate, the percentage of counts incorporated into DAG and TAG was increased at the expense of the counts incorporated into PLs. This effect was similar at 2.8 and 16.7 mmol/l glucose. These results indicate that glucose and palmitate have different effects on neutral lipid synthesis in islets upon prolonged exposure: glucose increases the total number of counts incorporated into complex lipids, whereas palmitate specifically directs partitioning of the counts toward neutral lipid synthesis.

Effects of palmitate and glucose on TAG content in islets. We next sought to determine whether prolonged and simultaneous exposure to palmitate and glucose leads to a net increase in TAG content. We first tested the modified assay against the previously published method using crude cellular extracts of islets(6,15,16,17,18,19)freshly isolated from normal Wistar, ZDF, and ZLC rats. The results are presented in Table 2 and indicate that the highest recovery of intracellular TAG is obtained using our modified method. TAG content was then measured in Wistar rat islets after a 72-h culture in the presence of 2.8 or 16.7 mmol/l glucose, with or without 0.5 mmol/l palmitate (Fig. 4). Palmitate did not affect TAG content in the presence of 2.8 mmol/l glucose(105 ± 5 ng/islet in the presence of palmitate vs. 76 ± 8 ng/islet in the absence of palmitate, n = 3, NS) but significantly increased it in the presence of 16.7 mmol/l glucose (303 ± 58 ng/islet in the presence of palmitate vs. 171 ± 31 ng/islet in the absence of palmitate, P = 0.03, n = 6). These results indicate that the presence of high glucose is necessary for prolonged exposure to palmitate to result in a net increase in TAG content in islets.

Concentration dependency of the effects of palmitate on neutral lipid metabolism and insulin mRNA levels in HIT-T15 cells. The concentration dependency of the effects of palmitate on neutral lipid synthesis, TAG content, and insulin mRNA levels was assessed in HIT-T15 cells(Fig. 5). First, HIT-T15 cells were cultured for 72 h in the presence of 11.1 mmol/l glucose, 5 μmol/l[U-¹⁴C]-labeled palmitate, and increasing concentrations of unlabeled palmitate at a constant palmitate:BSA molar ratio of 5:1. Analysis of incorporation of the counts into the various forms of complex lipids showed that palmitate dose-dependently shifted the partitioning of the counts into TAG and DAG, at the expense of PL, confirming the observations made in isolated islets (Fig. 5A). The effect of palmitate was statistically significant at 0.25 mmol/l and higher. Importantly, culturing HIT-T15 cells in the presence of increasing concentrations of BSA alone had no effect on the partitioning of the counts (data not shown). Next, HIT-T15 cells were cultured for 72 h in the presence of 11.1 mmol/l glucose with increasing concentrations of palmitate. TAG mass was measured as described above and was found to gradually augment as the concentration of palmitate was increased, with a significant and maximal effect observed at 0.5 mmol/l (P < 0.05, n = 5; Fig. 5B). This was associated with a dose-dependent decrease in insulin mRNA levels (Fig. 5Band C) that was statistically significant at and above 0.25 mmol/l. These results indicate that the effect of palmitate promoting neutral lipid synthesis in the presence of high glucose is concentration dependent and that there is an inverse relationship between TAG content and insulin gene expression.

Effects of palmitate, octanoate, palmitate methyl ester, and 2-bromopalmitate on insulin mRNA levels in HIT-T15 cells. We next investigated whether increased esterification into neutral lipids was necessary for the inhibitory effects of fatty acids on insulin mRNA levels(Fig. 6). After a 24-h exposure to 0.5 mmol/l palmitate, insulin mRNA was decreased by ∼40% compared with control in the presence of BSA alone, confirming our previous findings (3; Fig. 5). 2-Bromopalmitate,which irreversibly binds to carnitine-palmitoyl-transferase 1 (CPT-1) and inhibits longchain fatty acid β-oxidation(23), had a similar effect to palmitate. In contrast, the medium-chain fatty acid octanoate, which does not require CPT-1 to enter the mitochondria and is readily oxidized, and palmitate methyl ester, which is not activated into a fatty-acyl CoA in the cytosol, did not affect insulin mRNA levels. These results suggest that the fatty-acid—induced decrease in insulin mRNA levels is not related toβ-oxidation but requires activation of the fatty acid into a fatty-acyl CoA. Considering that fatty-acyl CoAs undergo either β-oxidation or esterification, these results indirectly suggest that activation of the esterification pathway is necessary for fatty acids to inhibit insulin gene expression.

This study was designed to assess whether the decrease in insulin gene expression upon prolonged exposure to palmitate in the presence of high glucose was associated with modulation in the expression of lipogenic and lipolytic enzymes in β-cells and with increased esterification of fatty acids into neutral lipids.

GPAT, DGAT, and HSL mRNA were detected, demonstrating for the first time that DGAT is expressed in pancreatic islets. In ZDF rat islets, the dramatic increase in TAG content is associated with a marked elevation in GPAT mRNA(6). However, we have not found GPAT mRNA to be increased in normal islets after prolonged exposure to palmitate. In other cell types, it has been suggested that DGAT, together with GPAT, is an important enzyme in TAG synthesis because it catalyzes the final and only committed step in this pathway(24). DGAT activity is increased in adipose tissue from obese Zucker rats(25). Its expression was not found to be modified in islets cultured in palmitate and glucose for 3 days. HSL, the rate-limiting enzyme for intracellular TG hydrolysis, is expressed and active in rat islets and several β-cell lines(22). Its expression, however,does not seem to be modified by prolonged exposure to either glucose or palmitate.

Pancreatic β-cells have the capacity to store energy in the form of TAG from both glucose and palmitate upon short-term incubation(26). Glucose is mainly incorporated into the glycerol-3-phosphate backbone of TAG(27). In the fasting state,the β-cell preferentially utilizes fatty acids as metabolic fuels, and therefore palmitate oxidation is high compared with esterification(28,29). At high glucose concentrations, lipid metabolism is switched to preferential incorporation into TAG and PL(28), at least in the short term. Our results show that these observations also apply to long-term situations and demonstrate that exposing islets to elevated concentrations of palmitate results in a sustained increase in intracellular TAG stores only if the concentration of glucose is elevated. The metabolic signals responsible for the switch from fatty-acid oxidation to esterification in the presence of high glucose in the β-cell are thought to be malonyl-CoA and cytosolic long-chain acyl-CoA (30). Acceleration of glucose metabolism following a rise in intracellular glucose concentration leads to increased cytosolic levels of malonyl-CoA(31,32),which potently inhibits CPT-1, thereby decreasing fatty-acid oxidation and resulting in accumulation of fatty acyl-CoA esters in the cytosol(33). Our results suggest that this hypothesis, originally proposed to explain the role of fatty acids in stimulus-secretion coupling in the β-cell, also applies to situations in which β-cells are chronically exposed to elevated levels of glucose and fatty acids. Furthermore, our results directly demonstrate that esterification of fatty acids into neutral lipids is increased upon prolonged exposure to glucose and palmitate, and that both nutrients have synergistic effects on neutral lipid metabolism. Glucose promotes incorporation of the fatty acid into complex lipids, whereas palmitate specifically directs partitioning toward neutral lipid synthesis. As expected, this results in a net increase in neutral lipid mass only when both fuels are present simultaneously. The fact that neutral lipid accumulation is not associated with changes in the level of expression of lipogenic and lipolytic enzymes suggests that activation of the esterification pathway is driven by the amount of substrates, i.e., fatty acids and glucose. Intracellular accumulation of neutral lipids results mainly from the balance between the activities of lipogenic and lipolytic enzymes. Our results indicate that overall lipogenic activity is augmented upon prolonged exposure to palmitate in the presence of high glucose, resulting in increased DAG and TAG content.

Various methods have been used for measuring TAG content in islets,generating quantitatively variable results. The differences between studies are in the mode of preparation of the lipid extract utilized for the assay. Most studies in ZDF animals have used crude cellular extracts and found intracellular levels of TAG to increase from <50 ng/islet in control rats to up to 1,000 ng/islet in 12-week-old ZDF animals(6,15,16,17,18,19). Two studies using normal islets in culture have extracted the lipids before the GPO Trinder assay(34,35). Our results are quantitatively similar to those of Zhou et al.(34) and uniquely demonstrate that the presence of glucose is necessary for TAG content to increase significantly upon prolonged exposure to exogenous palmitate.

In conclusion, the results presented herein demonstrate for the first time that neutral lipid synthesis is increased in normal β-cells exposed to elevated palmitate and glucose for prolonged periods of time. Glucose is required for palmitate to increase intracellular TAG mass, and both fuels have synergistic effects on neutral lipid metabolism. The glucose-dependent activation of the esterification pathway upon prolonged exposure to palmitate does not seem to be mediated by modulation of expression of key enzymes involved in neutral lipid metabolism, and it is therefore likely to be driven by the amount of available substrates.

I.B. was supported in part by the Association de Langue Française pour l'Etude du Diabète et des Maladies Métaboliques (ALFEDIAM) and the Aide aux Jeunes Diabétiques (AJD).

We thank Drs. Stewart A. Metz, Christopher J. Rhodes, and R. Paul Robertson for fruitful discussions and careful reading of the manuscript. We acknowledge the excellent secretarial assistance of Madeline Johnson.