Fatty acids rapidly induce the carnitine palmitoyltransferase I gene in the pancreatic beta-cell line INS-1.

Fatty acids are important metabolic substrates for the pancreatic β-cell, and long term exposure of pancreatic islets to elevated concentrations of fatty acids results in an alteration of glucose-induced insulin secretion. Previous work suggested that exaggerated fatty acid oxidation may be implicated in this process by a mechanism requiring changes in metabolic enzyme expression. We have therefore studied the regulation of carnitine palmitoyltransferase I (CPT I) gene expression by fatty acids in the pancreatic β-cell line INS-1 since this enzyme catalyzes the limiting step of fatty acid oxidation in various tissues. Palmitate, oleate, and linoleate (0.35 mM) elicited a 4-6-fold increase in CPT I mRNA. The effect was dose-dependent and was similar for saturated and unsaturated fatty acids. It was detectable after 1 h and reached a maximum after 3 h. The induction of CPT I mRNA by fatty acids did not require their oxidation, and 2-bromopalmitate, a nonoxidizable fatty acid, increased CPT I mRNA to the same extent as palmitate. The induction was not prevented by cycloheximide treatment of cells indicating that it was mediated by pre-existing transcription factors. Neither glucose nor pyruvate and various secretagogues had a significant effect except glutamine (7 mM) which slightly induced CPT I mRNA. The half-life of the CPT I transcript was unchanged by fatty acids, and nuclear run-on analysis showed a rapid (less than 45 min) and pronounced transcriptional activation of the CPT I gene by fatty acids. The increase in CPT I mRNA was followed by a 2-3-fold increase in CPT I enzymatic activity measured in isolated mitochondria. The increase in activity was time-dependent, detectable after 4 h, and close to maximal after 24 h. Fatty acid oxidation by INS-1 cells, measured at low glucose, was also 2-3-fold higher in cells cultured with fatty acid in comparison with control cells. Long term exposure of INS-1 cells to fatty acid was associated with elevated secretion of insulin at a low (5 mM) concentration of glucose and a decreased effect of higher glucose concentrations. It also resulted in a decreased oxidation of glucose. The results indicate that the CPT I gene is an early response gene induced by fatty acids at the transcriptional level in β- (INS-1) cells. It is suggested that exaggerated fatty acid oxidation caused by CPT-1 induction is implicated in the process whereby fatty acids alter glucose-induced insulin secretion.


centration of glucose and a decreased effect of higher glucose concentrations. It also resulted in a decreased oxidation of glucose. The results indicate that the CPT I gene is an early response gene induced by fatty acids at the transcriptional level in ␤-(INS-1) cells. It is suggested that exaggerated fatty acid oxidation caused by CPT-1 induction is implicated in the process whereby fatty acids alter glucose-induced insulin secretion.
Long chain fatty acids (LCFA) 1 exert divergent short and long term effects on pancreatic ␤-cell function. Acute administration of fatty acids potentiates glucose-induced insulin release both in vivo and in vitro (1)(2)(3)(4)(5)(6). Indirect evidence suggests that an increase in long chain fatty acyl-CoA (LC-CoA) is instrumental in this effect (reviewed in Refs. 6 and 7). In addition, studies in pancreatic islets show that palmitate increases cytosolic Ca 2ϩ in a fuel-dependent mechanism (8). In contrast, long term exposure of pancreatic islets to fatty acids causes an increased release of insulin at low (2-5 mM) concentrations of glucose and a decrease of glucose-induced insulin secretion, without changes in the response to non-nutrient secretagogues (9 -11). Short and long term effects of fatty acids may therefore be mediated by different mechanisms.
In vitro, the long term effect of LCFA on the ␤-cell requires at least 6 -24 h (10). It may therefore involve, in addition to a decrease in pyruvate dehydrogenase activity (12), changes in the expression of key enzymes of glucose and/or fatty acid metabolism. Consistent with this possibility, it is now well established that calorigenic nutrients are not only metabolic substrates and/or short term regulators of enzyme activities. Glucose and fatty acids modulate the expression level of a number of genes encoding metabolic enzymes in a variety of cell types (13,14). Fatty acids have been reported to induce several genes of hepatic fatty acid metabolism, including those encoding fatty acid-binding protein (15) and enzymes of peroxisomal (16) and mitochondrial (16 -18) ␤-oxidation. Because excessive lipid levels and their oxidation appear to play a major role in the development of insulin resistance and non-insulindependent diabetes (19,20), identification of target genes of fatty acids in the pancreatic ␤-cell may help in the determination of candidate genes potentially responsible for alterations in fuel metabolism and insulin secretion.
Carnitine palmitoyltransferase I (CPT I) is considered as the rate-limiting enzyme regulating fatty acid oxidation in mito-* This work was supported in part by a fund from the Faculty of Medicine of the University of Geneva and by Swiss National Science Foundation Grant 3200-045957.95/1, grants from the Canadian Diabetes Association, the Medical Research Council of Canada, National Institutes of Health Grant DK18573, and a Juvenile Diabetes Foundation International/NIH Interdisciplinary Program Project initiative. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 (21). As such, it plays a central role in the partitioning of fatty acids between mitochondrial oxidation and their accumulation as LC-CoA and/or complex lipids in the cytoplasm (reviewed in Ref. 22). CPT I is inhibited by malonyl-CoA, and this inhibition is overcome by increasing LC-CoA (22). The activity of acetyl-CoA carboxylase, the provider of malonyl-CoA, is therefore also of critical importance in the regulation of fatty acid oxidation (21). In addition to changes in the concentrations of malonyl-CoA and variations in the sensitivity of the CPT I enzyme to malonyl-CoA inhibition, the amount of fatty acid oxidized may also be regulated by changes in the maximal activity of CPT I (21). For example, cAMP and fatty acids increase the expression level of the CPT I gene in cultured fetal rat hepatocytes (23).
The malonyl-CoA/CPT I interaction has emerged as a key component of a fuel "cross-talk" metabolic signaling system in a number of tissues including the liver (24), heart (25), and skeletal muscle (26). In addition, CPT I is a potentially important site of pharmacological intervention in diabetes where fatty acid oxidation is excessive and impairs glucose homeostasis (20). With respect to the ␤-cell, the hypothesis has been proposed that malonyl-CoA, via its inhibitory action on CPT I and a resulting rise in cytosolic LC-CoA, is implicated in conjunction with the K ATP channel pathway in the transduction mechanisms whereby nutrients induce the insulin secretory process (7,27,28).
To better understand how exposure of the ␤-cell to LCFA modifies fuel metabolism and insulin secretion and to gain insight into CPT I gene regulation, the effect of fatty acids and other nutrients on the expression of the CPT I gene was studied in the pancreatic ␤-cell line INS-1. The data show that LCFA rapidly induce transcription of the CPT I gene which in turn leads to an increased activity of CPT I and to a higher capacity of INS-1 cells to oxidize fatty acids.

EXPERIMENTAL PROCEDURES
Cell Culture and Incubation-INS-1 cells (at passages below 85) were grown in monolayer cultures as described previously (29) in RPMI 1640 medium containing 11 mM glucose supplemented with 10 mM HEPES, 10% heat-inactivated fetal calf serum, 2 mM L-glutamine, 1 mM sodium pyruvate, 50 M ␤-mercaptoethanol, 100 IU/ml penicillin, and 100 g/ml streptomycin in an humidified atmosphere (5% CO 2 , 95% air). After 6 -7 days (60 -70% confluency), they were incubated in the same medium with various fatty acids or test substances and used for measurement of CPT I or actin mRNA levels, CPT I activity, fatty acid or glucose oxidation, and insulin secretion. Albumin-bound fatty acids were prepared by stirring fatty acid sodium salts at 45°C with defatted bovine serum albumin. After adjustment of the pH to 7.4, the solution was filtered through a 0.22-m filter, and the fatty acid concentration was measured using a NEFA PAP kit (bioMerieux, Lyon, France), except for 2-bromopalmitate which could not be measured by this method. For incubations longer than 24 h, the medium was changed every day to maintain a constant concentration of fatty acid. Unless otherwise stated, the final concentration of BSA in the culture medium was 0.5%.
Insulin Secretion-INS-1 cells were plated (10 5 cells/well) into 24well plates and cultured for 7 days in RPMI medium. They were then cultured for 3 days in the same medium containing 5 mM glucose and 0.5% BSA or BSA-bound fatty acids (0.35 mM) (30). They were then washed and preincubated for 30 min in Krebs-Ringer bicarbonate buffer containing 5 mM glucose, 10 mM HEPES, and 0.5% BSA. At this point, the glucose concentration was raised for 11 or 25 mM, and the insulin concentration in the medium was determined by radioimmunoassay using rat insulin as standard (30). Total cellular insulin content was measured after acid/ethanol (1.5% HCl, 75% ethanol) extraction (30).
CPT I mRNA Analysis-Total RNA was extracted from cells by the guanidium isothiocyanate method (31). RNA samples (15 g) were denatured in formamide and formaldehyde at 95°C for 3 min. Northern blot analyses were performed after 1% agarose-gel electrophoresis in 2.2 M formaldehyde (32). After transfer to nylon membranes, the filters were hybridized with rat liver CPT I (33) or ␤-actin (34) cDNA probes labeled with [␣-32 P]dCTP using the Redyprime labeling system kit. The autoradiograms were analyzed by densitometer scanning.
In Vitro Transcription Assay-Nuclei isolation and nuclear run-on transcription assays were performed according to Ref. 35. Briefly, nascent transcripts were elongated in vitro in the presence of [ 32 P]UTP and 2.1 mg/ml heparin. The [ 32 P]RNAs obtained were subjected to mild alkaline hydrolysis (30 min, 50°C, 50 mM Na 2 CO 3 ) and hybridized to 4 g/dot of CPT I cDNA immobilized on nitrocellulose membranes.
Isolation of Mitochondria and Measurement of CPT I Activity-Cells were cultured for various periods of time with supplemented RPMI 1640 medium containing 0.5% BSA or BSA-oleate (0.4 mM final concentration). For each time point, cells were scraped from five 115-mm diameter Petri dishes and washed twice with cold phosphate-buffered saline. Mitochondria were isolated by differential centrifugation (36), followed by purification on a Percoll gradient (37), and used for the assay of CPT I activity (36).
Measurement of Palmitate and Glucose Oxidation-INS-1 cells were cultured with supplemented RPMI 1640 medium containing 0.5% BSA or BSA-palmitate (0.35 mM final concentration) for 3 days. They were then trypsinized and preincubated as a suspension in siliconized tubes for 1 h at 37°C in RPMI 1640 medium containing 5 mM glucose and 10 mM HEPES without other supplementation. After centrifugation and resuspension of the cells in the same medium with varying glucose concentrations, palmitate oxidation was measured as 14 CO 2 production from [U-14 C]palmitate bound to BSA. Cells were incubated at 37°C for 2 h in the presence of [U-14 C]palmitate (0.35 mM, 1 Ci/mol), and the reaction was stopped by the addition of 0.2 ml of 10% trichloroacetic acid. Benzethonium hydroxide (0.3 ml) was injected into the small wells suspended to the rubber caps of the tubes, and, after 4 h at room temperature, the trapped 14 CO 2 was measured by liquid scintillation counting. The recovery of 14 CO 2 , as assessed with NaH 14 CO 3 , was 74 Ϯ 5% (n ϭ 3). Cells were counted in a Coulter Counter, and the data were expressed as nanomoles of palmitate oxidized/2 h ϫ 10 6 cells.
Glucose oxidation was measured as 14 CO 2 production from [U-14 C]glucose (0.1 Ci/mol) using the same experimental design as for the measurement of palmitate oxidation, except that cells were incubated for 60 min with [U-14 C]glucose in the absence of fatty acids. The data are expressed as micromoles of glucose oxidized/h ϫ 10 6 cells.
Statistical Analysis-All results are expressed as means Ϯ S.E. Statistical significance was calculated with the Student's t test. Fig. 1, the lag time of the induction was shorter than 60 min, and a 2-fold increase in the expression level of CPT I mRNA occurred at 1 h. A maximal effect was observed after 3 h, with a 6-fold accumulation of CPT I mRNA. The data were similar when related to the actin transcript since this parameter did not vary with time (data not shown, see also Fig. 2). The nonmetabolizable analog, 2-bromopalmitate, was also tested. It caused a similar induction of CPT I mRNA up to 6 h. However, the level of CPT I transcript remained constant for at least 24 h in the presence of 2-bromopalmitate, whereas it declined progressively with time in the presence of palmitate. The difference between the action of the two fatty acids is most likely due to palmitate metabolism by the cells which presumably caused a time-dependent fall in the medium concentration of the fatty acid. Thus, in another series of experiments we measured palmitate in the medium at time zero and following a 10-h incubation period. After 10 h, the palmitate concentration was 68 Ϯ 7% (mean Ϯ S.E., n ϭ 3) of that present at time zero. Since the action of the tested fatty acids was maximal between 3 and 6 h, all subsequent CPT I mRNA measurements were carried out at the 6-h time point unless otherwise stated.

Effect of Various Fatty Acids and Other Nutrients on CPT I mRNA Accumulation-Palmitate caused a marked induction of the CPT I transcript in INS-1 cells. As shown in
The dose dependence of the effect of palmitate and oleate is presented in Fig. 2. For this experiment, a concentration of 1% BSA was used to allow testing of a high (0.6 mM) concentration of palmitate while maintaining the molar ratio of fatty acid to BSA below 7 (5,38). At 0.1 mM, both palmitate and oleate significantly (p Ͻ 0.05) increased CPT I mRNA. A similar dose dependence and maximal effect was observed with the two fatty acids (Fig. 2).
To determine whether the degree of unsaturation of the fatty acids influences CPT I gene induction, we tested the actions of the three most abundant circulating fatty acids, i.e. palmitate (C16:0), oleate (C18:1), and linoleate (C18:2) in the same experiment. All fatty acids at a concentration of 0.35 mM induced CPT I mRNA to approximately the same extent and similarly to 2-bromopalmitate ( Fig. 3), suggesting that the inductive effect is not dependent on the type of long chain fatty acid.
To assess whether the action of the fatty acids requires metabolism beyond their activation to long chain fatty acyl-CoA derivatives, we tested the effect of the fatty acid oxidation and CPT I inhibitor, etomoxir. This compound did not modify the effect of oleate (Fig. 3). Etomoxir alone had no effect on CPT I mRNA (data not shown). It should be mentioned that etomoxir (20 M) inhibited fatty acid oxidation of INS-1 cells by only 25-30% (data not shown). Within the limit of the relatively weak inhibitory action of this drug on fatty acid oxidation in INS-1 cells, the data are consistent with the view that the mitochondrial oxidation of fatty acids is not required for the action of this class of nutrient on CPT I gene induction. A stronger argument in favor of this view is the observation reported in Figs. 1 and 3 that 2-bromopalmitate, which is activated to 2-bromopalmitoyl-CoA but is not further metabolized (22), induced the CPT I transcript to a similar extent as did all the tested metabolizable fatty acids.
Since the inductive effect was extremely rapid (Fig. 1), we asked whether the CPT I gene is an early response gene induced by activation of pre-existing factors (39). Accordingly, INS-1 cells were stimulated with oleate in the presence of cycloheximide. Cycloheximide alone had no effect. The action of the fatty acid was not inhibited by the protein synthesis inhibitor; rather it was enhanced (p Ͻ 0.02). This is a characteristic feature of several early response genes such as c-fos that show a superinduction in response to various agonists in the presence of protein synthesis inhibitors, possibly because the transcript is degraded at a reduced rate (39). Thus, the CPT I gene behaves in INS-1 cells as an early-response gene whose induction does not require de novo protein synthesis.
The effect on CPT I mRNA was stimulus-specific and restricted to fatty acids; other nutrients, in particular glucose (20 mM) and pyruvate (10 mM), had no effect (data not shown). A significant (p Ͻ 0.05) 2.5 Ϯ 0.7-(n ϭ 3) fold increase in CPT I mRNA was measured with 7 mM glutamine. The Ca 2ϩ and cAMP transduction systems do not mediate the action of the fatty acid since elevated K ϩ , which promotes Ca 2ϩ influx (40), and the adenylate cyclase activator, forskolin, were ineffective (data not shown). Likewise, the protein kinase C system appears not to be implicated since the protein kinase C activator, phorbol 12-myristate 13-acetate (PMA), had no effect (data no shown). In addition, down-regulation of protein kinase C by a 24-h preincubation period with PMA (10 Ϫ7 M) did not modify the induction of CPT I mRNA by palmitate (6.3 Ϯ 0.4-and 5.7 Ϯ 0.4-fold in control and protein kinase C down-regulated cells, respectively (means Ϯ S.E. of 3 experiments).
Palmitate Induction of the CPT I Gene Occurs at the Transcriptional Level-Fatty acid-induced accumulation of CPT I mRNA might result from alterations in transcription rate and/or mRNA turnover. Run-on assays were carried out using nuclei from INS-1 cells incubated for 45 min with palmitate. The results in Fig. 4 clearly show that palmitate increased dramatically and rapidly the CPT I gene transcriptional rate. Thus, in the absence of fatty acid, the transcription of the gene was undetectable, whereas a strong signal was observed in nuclei obtained from fatty acid-treated cells. Under the same experimental condition, palmitate did not affect the transcriptional rates of the glyceraldehyde-3-phosphate dehydrogenase and the 18 S ribosomal genes. To assess whether palmitate modifies CPT I mRNA stability, the half-life of CPT I mRNA was determined in the presence of the transcription inhibitor actinomycin D. The measured half-life of the CPT I transcript was short for one encoding a metabolic enzyme (about 3 h) and was unchanged by fatty acids (3.5 Ϯ 0.5 h and 2.5 Ϯ 0.4 h with BSA or palmitate, respectively, not significant). Thus, transcriptional activation of the CPT I gene by palmitate appears to fully account for CPT I mRNA induction by the fatty acid.
Long Term Exposure of INS-1-Cells to Fatty Acids Increases CPT I Enzyme Activity, Fatty Acid Oxidation, and Impairs Glucose Oxidation and Glucose-induced Insulin Release-Fatty acid-induced CPT I mRNA accumulation was followed by a time-dependent increase of CPT I activity of isolated mitochondria after incubation with 0.4 mM oleate (Fig. 5). CPT I activity increased by 40% after 4 h and attained 2-3 times the initial value between 24 and 72 h.
The effect of fatty acids on the capacity of INS-1 cells to oxidize fatty acids was also measured after 3 days of incubation with 0.35 mM palmitate or BSA. The data of Table I show that exposure to fatty acids increased palmitate oxidation at all tested glucose concentrations (1 to 11 mM). As expected from other studies (27,41), increasing the glucose concentration decreased palmitate oxidation. It is noteworthy that both CPT I activity and fatty acid oxidation were increased by pre-exposure to fatty acids to a similar extent (about 2.5-fold) at low concentrations of glucose. This is in accordance with the concept that CPT I gene induction is causally implicated in the observed exaggerated fatty acid oxidation.
INS-1 cells cultured for 3 days with 0.35 mM palmitate or oleate showed an increased "basal" (glucose 5 mM) insulin secretion and a blunted glucose-stimulated insulin secretion (Table II). Glucose oxidation, measured in the absence of exogenous fatty acids, was also decreased in cells cultured for 3 days with palmitate in comparison with control cells (Table III). The data indicate that fatty acids produce changes in glucose-induced insulin secretion and glucose oxidation in INS-1 cells similar to those observed in normal pancreatic islets (10,11). Hence, INS-1 cells at relatively low passages are appropriate for studies aimed at understanding the mechanism whereby fatty acids affect ␤-cell function. DISCUSSION The present report shows that LCFA, at concentrations within the physiological range (0.1-0.6 mM), are major regulators of the CPT I gene in clonal pancreatic ␤-(INS-1) cells. Other nutrients including glucose, pyruvate, and glutamine have little or no effect. This indicates that the inductive process is specific for fatty acids. The action of LCFA is rapid since CPT I mRNA is induced within less than 60 min and a 2-fold rise in the transcript occurs 1 h following LCFA addition to cells. It is also quantitatively important as LCFA caused a 5-7-fold induction of CPT I mRNA at their maximal effective concentration. The three most abundant circulating LCFA (the saturated palmitate, monounsaturated oleate, and polyunsaturated linoleate) all induced CPT I mRNA to the same extent. This suggests that the gene regulation process is not specific for a given class of fatty acid as it is for lipogenic enzymes in the liver where polyunsaturated fatty acids only reduced the expression of acetyl-CoA carboxylase mRNA (14).
Transcriptional activation of the CPT I gene appears to account fully for CPT I mRNA accumulation since a pronounced increase in the rate of transcription of the gene was detected 45 min after LCFA addition, and LCFA did not significantly change CPT I mRNA stability. The induction of CPT I mRNA by LCFA is apparently an early and direct event since de novo protein synthesis inhibition by cycloheximide did not suppress the increased expression of CPT I mRNA. It thus appears that the CPT I gene is an early response gene (39) induced by fatty acids in pancreatic ␤-(INS-1) cells.
The mechanism of CPT I gene induction by LCFA remains to be elucidated. LCFA are known to increase ␤-cell cytosolic Ca 2ϩ (8) and LCFA, and their CoA derivatives activate protein kinase C (PKC) isoenzymes (28,42). However, the Ca 2ϩ signaling system is not implicated since elevated K ϩ which promotes Ca 2ϩ influx and secretion in normal ␤-cells and INS-1 cells (40) did not alter CPT I mRNA level. An effect of PKC is unlikely for two reasons. Unsaturated fatty acids are much more effective than saturated LCFA in activating protein kinases C (42), but both types of fatty acid-induced CPT I gene expression to a similar extent. Moreover, PMA had no effect, and the increase of CPT I mRNA caused by palmitate was unchanged after down-regulation of PKC by a 24-h pre-exposure to PMA, a procedure known to efficiently reduce PKC enzymes in the ␤-cell (43). It is noteworthy that CPT I gene induction does not require metabolism of LCFA beyond long chain fatty acyl-CoA formation since 2-bromopalmitate, which is readily converted into a nonmetabolizable CoA ester, was equally effective as naturally occurring LCFA. Thus, it appears that LCFA directly or in their activated CoA form mediate CPT I gene regulation. Importantly, LC-CoA modulates the activity of many enzymes with little distinction between saturated and unsaturated forms (44) and have been implicated in the regulation of a number of lipid metabolism genes in bacteria (45). The small inductive effect of glutamine might have been caused by the cytosolic accumulation of LCFA or LC-CoA in the amino acidtreated cells because glutamine is a very potent inhibitor of fatty acid oxidation in ␤-cells (5,46).
LCFA are known to influence gene transcription of a number of genes through PPARs which are nuclear receptors closely related to the steroid-thyroid hormone superfamily (16). To date, three types of PPAR have been described (␣, ␤, and ␥). The target genes of PPARs encode enzymes of lipid metabolism and homeostasis (16). PPARs are also involved in differentiation processes, in particular those of adipocytes (47). PPARs heterodimerize with the retinoic acid X receptor and alter the transcription of target genes after binding to response elements consisting of a direct repeat of a nuclear receptor hexameric DNA recognition motif spaced by one nucleotide (DR-1) (16,47). Functional peroxisome poliferator response elements have been identified in the regulatory region of numerous genes encoding enzymes involved in lipid metabolism (16,47). PPAR genes are differentially expressed in a wide range of tissues (48). A study of the expression pattern of PPAR in rat islets has shown that PPAR␤ is predominantly expressed, whereas the levels of PPAR␣ and -␥ are low (48). INS-1 cells show a similar expression pattern of PPARs as indicated by reverse transcription-polymerase chain reaction. 2 The predominance of PPAR␤ in INS-1 cells may explain why neither PGJ2, which acts through PPAR␥ (49), nor Wy-14643, acting via PPAR␣ (49), had no effect on CPT I gene induction in this cell type (not shown). Thus, PPAR␤ for which there is no known specific agonist, may mediate LCFA induction of the CPT I gene. However, other mechanisms (independent of PPAR) should be considered, as suggested in studies with cultured hepatocytes (23). After 1-3 days of exposure to elevated LCFA, the induction of the CPT I gene was associated with a 2.5-fold increase in CPT I enzymatic activity, as assessed in isolated mitochondria from INS-1 cells. Palmitate oxidation was also enhanced about 2.5fold, suggesting that increased CPT-I activity is causally implicated in the enhanced fatty acid oxidation in the LCFAtreated cells. The parallelism between the enhancement of CPT I activity and palmitate oxidation is consistent with the idea that CPT I is limiting for LCFA oxidation in ␤-(INS-1) cells, as is thought to be the case in other cell types (21). With respect to insulin secretion, LCFA treatment of INS-1 cells reproduces precisely what has been described in rat (10,11,50) and human (51) islets both in vivo and in vitro; i.e. it enhances secretion at low concentrations of glucose while suppressing the normal response to higher (larger than 5 mM) concentrations of the sugar.
The action of LCFA on the CPT I gene is potentially relevant 2 N. Voilley and M. Prentki, unpublished observations.  to ␤-cell physiopathology not only because the malonyl-CoA/ CPT I interaction may be implicated in the mechanism whereby nutrients promote the release of insulin (6,7,28), but also because of the inverse relationship between the oxidation of fatty acids and glucose in various tissues including pancreatic islets (20,52,53). LCFA are major substrates for islet metabolism at low concentrations of glucose, particularly in the fasted state (53). The interactive regulation of pyruvate dehydrogenase and CPT I is an established feature of the relationship between carbohydrate and fatty acid metabolism since activation of either system generates regulatory metabolites that suppress the activity of the other (52). Thus, CPT I induction by LCFA could be instrumental in the establishment of a Randle cycle (54) with reduced pyruvate dehydrogenase activity and glucose oxidation as described in rat islets following a 48-h exposure to elevated LCFA (10,55). Consistent with this view, the fatty acid oxidation inhibitor, etomoxir, partially reversed these defects caused by LCFA (10,51). Both CPT I induction and the down-regulation of acetyl-CoA carboxylase by LCFA 3 with a possible reduction in malonyl-CoA formation could participate in the molecular defect involved in the lack of sensitivity of the ␤-cell to elevated concentrations of glucose. However, the complexity of the action of LCFA on the ␤-cell is emphasized by the ability of these substrates to cause exaggerated secretion of insulin at low glucose. There are some reasons to believe that the increased expression of a low K m hexokinase is implicated in the latter phenomenon (56). An elevated concentration of cytosolic LC-CoA in fatty acid-treated cells (5), which is thought to be implicated in insulin secretion (6,7,28), is an alternative possibility.
In conclusion, the CPT I gene is an early response gene regulated at the transcriptional level by long chain fatty acids in ␤-(INS-1) cells. The rapidity of CPT I gene induction is of potential interest for ␤-cell metabolism and the regulation of genes encoding enzymes of glucose and lipid metabolism. Thus, accelerated fatty acid oxidation is expected to change carbohydrate metabolism and LC-CoA partitioning between the mitochondria, cytosolic, and possibly nuclear compartments. Experiments are underway to explore the possibility that CPT I gene induction per se may play an important role in causing some of the pleiotropic changes in metabolism, secretion, and late gene expression in ␤-cells exposed for long periods of time to elevated LCFA.