Clinical Care/Education/Nutrition/Psychosocial Research ORIGINAL ARTICLE

Effect of Oral Sebacic Acid on Postprandial Glycemia, Insulinemia, and Glucose Rate of Appearance in Type 2 Diabetes

1 1 AMERIGO IACONELLI, MD ANGELA FAVUZZI, MD beneficial in terms of reduction of circu- 2 3 AMALIA GASTALDELLI, PHD CHRISTOPHE BINNERT, PHD lating levels of glucose, insulin, and free 1 3 CHIARA CHIELLINI, PHD KATHERINE MACE´, PHD 1 1 fatty acids in type 2 diabetes. Thus, the DONATELLA GNIULI, MD GELTRUDE MINGRONE, MD, PHD availability of snacks poor in fat and that do not induce hyperglycemia and/or overstimulate insulin secretion would be OBJECTIVE — Dicarboxylic acids are natural products with the potential of being an alter- a good tool in the diet of insulin-resistant, nate dietary source of energy. We aimed to evaluate the effect of sebacic acid (a 10-carbon type 2 diabetic subjects. dicarboxylic acid; C10) ingestion on postprandial glycemia and glucose rate of appearance (Ra) in healthy and type 2 diabetic subjects. Furthermore, the effect of C10 on insulin-mediated Dicarboxylic acids are naturally oc- glucose uptake and on GLUT4 expression was assessed in L6 muscle cells in vitro. curring substances produced by both higher plants and animals via ␻-oxidation RESEARCH DESIGN AND METHODS — Subjects ingested a mixed meal (50% car- of fatty acids (6,7). In plants, long-chain bohydrates, 15% proteins, and 35% lipids) containing 0 g (control) or 10 g C10 in addition to dicarboxylic acids are components of nat- the meal or 23 g C10 as a substitute of fats. ural protective polymers, cutin and RESULTS — In type 2 diabetic subjects, the incremental glucose area under the curve (AUC) suberin, which support biopolyesters in- decreased by 42% (P Ͻ 0.05) and 70% (P Ͻ 0.05) in the 10 g C10 and 23 g C10 groups, volved in waterproofing the leaves and respectively. At the largest amounts used, C10 reduced the glucose AUC in healthy volunteers fruits, regulating the flow of nutrients also. When fats were substituted with 23 g C10, AUC of Ra was significantly reduced on the order among various plant cells and organs, and of 18% (P Ͻ 0.05) in both healthy and diabetic subjects. The insulin-dependent glucose uptake minimizing the deleterious impact of Ϯ Ϯ ϭ by L6 cells was increased in the presence of C10 (38.7 10.3 vs. 11.4 5.4%; P 0.026). This pathogens (7). In animals and humans, increase was associated with a 1.7-fold raise of GLUT4. medium chain dicarboxylic acids, which CONCLUSIONS — Sebacic acid significantly reduced hyperglycemia after a meal in type 2 include prevalently sebacic (C10) and do- decanedioic (C12) acids, derive from the diabetic subjects. This beneficial effect was associated with a reduction in glucose Ra, probably due to lowered hepatic glucose output and increased peripheral glucose disposal. ␤-oxidation of longer chain dicarboxylic acids (8). Long-chain dicarboxylic acids, Diabetes Care 33:2327–2332, 2010 in turn, are formed from the correspon- dent fatty acids by ␻-oxidation in the mi- he World Health Report launched in sis of glycogen of ϳ25–45% compared crosomal membranes (9) or are taken in 2002 by the World Health Organi- with that in nondiabetic subjects (2). In- with a diet rich in vegetables (7). T zation advised that more than 1 bil- creased hepatic gluconeogenesis has been We have shown previously that me- lion adults worldwide are overweight and considered to be responsible for elevated dium-chain dicarboxylic acids represent a at least 300 million are clinically obese. hepatic glucose output in type 2 diabetes suitable alternate energy substrate to glu- Type 2 diabetes can be considered a (3). When glycogen is available in ade- cose in clinical conditions with marked threatening obesity-related disease be- quate amounts, there is an autolimitation insulin resistance and/or impaired aero- cause hyperglycemia causes relevant of hepatic glucose production. In diabe- bic glycolysis (10). Interestingly, in type 2 complications such as micro- and mac- tes, a breakdown of this autoregulation diabetes, medium-chain dicarboxylic ac- roangiopathy. Patients with type 2 diabe- may occur if glycogenolysis is limited by ids are rapidly oxidized, do not stimulate tes exhibit increased hepatic glucose glycogen depletion (4). insulin secretion, and reduce muscle fa- output, which is identified as the main Jenkins et al. (5) have shown that tigue (11). Nevertheless, the effect of C10 cause of fasting hyperglycemia and is as- spreading the nutrient load over a longer or C12, not as a substitute but in addition sociated with impaired plasma glucose period of time by increased meal fre- clearance (1) and reduced hepatic synthe- quency, the so-called nibbling diet, is to available carbohydrates, on glucose ho- meostasis has never been studied. ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● On this basis, our aim was to investi- From the 1Department of Internal Medicine, Catholic University of Rome, Rome, Italy; the 2Institute of gate the effect of oral administration of Clinical Physiology, Consiglio Nazionale delle Ricerche Pisa, Pisa, Italy; and the 3Nestle´ Research Centre, C10 on postprandial glycemia, insuline- Lausanne, Switzerland. mia, and glucose rate of appearance (Ra) Corresponding author: Geltrude Mingrone, [email protected]. Received 9 April 2010 and accepted 6 August 2010. Published ahead of print at http://care.diabetesjournals. in type 2 diabetic subjects compared with org on 19 August 2010. DOI: 10.2337/dc10-0663. that in healthy volunteers. To further elu- © 2010 by the American Diabetes Association. Readers may use this article as long as the work is properly cidate the mechanism of action of sebacic cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. acid in diabetes, insulin-mediated glucose org/licenses/by-nc-nd/3.0/ for details. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby uptake and GLUT4 protein expression marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. were assessed in L6 cells in vitro. care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 11, NOVEMBER 2010 2327 Oral sebacic acid in type 2 diabetes

RESEARCH DESIGN AND Table 1—Mixed meal composition METHODS — From October 2006 to September 2007, 10 obese type 2 diabetic Carbohydrates Proteins Lipids Sebacic acid subjects and 10 healthy volunteers were enrolled. Women in fertile age were asked Control (450 kcal) 50 15 35 0 to avoid pregnancy during the study pro- 10 g C10 (450 kcal ϩ 60 kcal from C10 44 13 31 12 tocol. In addition, before starting each ex- 23 g C10 (450 kcal) 50 15 0 35 perimental session a pregnancy test was Data are kcal%. performed, and pregnant women were excluded from the investigation. All women were studied in the follicular Blood samples for the determination monitored, and enrichment was calcu- phase of their menstrual cycle. None of of 6,6-D-glucose enrichment, plasma C- lated as the ratio of 202 to 200. C10 anal- the subjects previously or currently had peptide, and plasma insulin concentra- ysis was performed as described major endocrinological, renal, cardiac, tions were obtained before the tracer previously (14). respiratory, liver, or gastrointestinal infusion was started and every 10 min disorders. during the last 20 min of the 2 h preced- Calculations All subjects underwent a 75-g oral ing the meal ingestion (equilibration pe- A physiological and isotopic steady state glucose tolerance test to measure the Mat- riod). Starting from the meal ingestion was achieved during the last 20 min of the suda index, as an index of insulin resis- time, plasma samples were obtained every basal period and of the meal period; tance (12). None of the diabetic subjects 20 min for 5 h. Therefore, overall the therefore, the glucose rate of appearance were taking oral hypoglycemic agents or study duration was7hor420min. (Ra), which is equivalent to endogenous were diagnosed with type 2 diabetes 2–5 Each subject voided to empty his/her glucose production, and rate of disap- years before the study and their A1C was bladder before starting the experimental pearance (Rd) were calculated as the 5.5–7.5%. protocol. The subsequent 24-h urine tracer infusion rate divided by the tracer- The protocol was approved by the samples were collected in a container with to-tracee ratio. During the meal, glucose ethics committee of the Catholic Univer- 0.1% sodium azide to prevent bacterial non–steady-state conditions prevail, and, sity, School of Medicine, in Rome, Italy. growth. thus, Ra and Rd were calculated using the All of the subjects signed a written in- Steele equation (15). Glucose flux rates formed consent form before starting the Body composition are expressed per kilogram of body study. After subjects voided and were weighed, weight. Glucose clearance was calculated body composition was assessed by dual- as Rd divided by plasma glucose concen- Chemicals x-ray absorptiometry (Lunar Prodigy; GE tration. AUCs of glucose and insulin con- Sebacic acid was purchased from Sigma- Lunar, Madison, WI) to measure fat-free centrations and of glucose Ra and Aldrich (St. Louis, MO) and purified by mass, fat mass, and percentage of fat mass. clearance were calculated using the trap- Real s.r.l. (Como, Italy). It was pyrogen- ezoidal method. AUC of meal glucose free and contaminant-free with a degree Analytical methods clearance was normalized to AUC of insu- of purification (gas-liquid chromatogra- Plasma insulin and C-peptide were mea- lin to account for different insulin con- phy/mass spectrometry) of 99.5%. 6,6-D- sured in duplicate by double-antibody ra- centrations observed during the three Glucose was purchased from Mass Trace dioimmunoassays (Linco Research, St. tests. Insulin secretion rate was computed (Woburn, MA). Charles, MO). Intra-assay variation was by C-peptide deconvolution (16). 4.9 and 2.4%, and inter-assay variation Experimental protocol was 5.9 and 7.1%, respectively. Plasma Cell culture and glucose uptake The subjects were studied on three sepa- glucose was monitored immediately after L6 myoblasts were grown and maintained rate occasions in a randomized way at a blood withdrawal with an Analox GM9 in ␣-minimal essential medium contain- distance of 1 month. After an overnight glucose-analyzer (Beckman Instruments, ing 2% FBS and 1% antibiotic-antimy- fast, during which subjects were in- Fullerton, CA). cotic mixture in an atmosphere of 5% structed not to drink, the study was initi- Plasma glucose enrichment due to CO2 at 37°C as described elsewhere ated at 8:00 A.M. An indwelling catheter 6,6-D-glucose was determined by gas (17,18). The cells were cultivated over- was placed into an antecubital vein for chromatography-mass spectrometry on a night in 12-well plates with or without isotope infusion. A second catheter was gas chromatograph 5890/mass spectrom- 0.2 mmol/l albumin-bound C10 and se- inserted retrogradely into a wrist vein of eter 5972 (Hewlett Packard, Palo Alto, rum starved for 4 h before being incu- the ipsilateral hand, and the hand was CA) equipped with a 30-m capillary col- bated for 20 min with 10 nmol/l insulin. placed into a heated box (60°C) to achieve umn, as described by Wolfe (13). In brief, Subsequently, myoblasts were washed arterialization of venous blood. Basal rates 100 ␮l of plasma was deproteinized using twice, and glucose transport was assayed of glucose turnover were assessed after 2 h 2 ml of cold methanol, and the superna- in HEPES-buffered saline solution (140 of an adjusted primed (22 ␮mol/kg), con- tant was dried and derivatized using ace- mmol/l NaCl, 20 mmol/l HEPES-sodium, ␮ Ϫ1 Ϫ1 tinuous (0.22 mol kg min ) infu- tic anhydride-pyridine (1:1) to form 2.5 mmol/l MgSO4, 1 mmol/l CaCl2, and ␮ sion of 6,6-D2-glucose (Cambridge penta-acetate glucose. The samples were 5 mmol/l KCl, pH 7.4) containing 10 M Ϫ1 Isotope Laboratory, Boston, MA) (13). then dried again and dissolved with ethyl 2-deoxy-D-glucose (0.5 ␮Ci ml 2-de- 3 Two hours after the beginning of the ses- acetate for injection into the gas chro- oxy-D-[ H]glucose). The incubation me- sion, the subjects ingested a mixed meal matograph-mass spectrometer. The frag- dium was aspirated, the cells were washed (see Table 1 for meal composition). ments 200, 201, and 202 were with ice-cold saline, and 1 ml of NaOH

2328 DIABETES CARE, VOLUME 33, NUMBER 11, NOVEMBER 2010 care.diabetesjournals.org Iaconelli and Associates

(0.05 mol/l) was added to each well. Cell lysates were transferred to scintillation vi- als for radioactivity counting. Nonspecific uptake was determined in the presence of cytochalasin B (10 ␮M) and was sub- tracted from all values.

Immunoprecipitation and Western blot To measure the total GLUT content, cell lysates were subjected to 10% SDS-PAGE and Western blotting (19). Filters were incubated with GLUT4 antibodies (Ab- nova, Prodotti Gianni, Milano, Italy) for 14 h at 4°C and then with peroxidated anti-antibodies (Amersham Biosciences, Cologno Monzese, Italy) for1hatroom temperature. GLUT4 was finally revealed by detection of chemoluminescence with autoradiography.

Statistical analysis Data are presented as means Ϯ SD, unless otherwise specified. For the clinical study differences among the three sessions were analyzed by nonparametric (Wilcoxon’s signed-rank test) tests for paired samples corrected for Bonferroni inference. Anal- yses were performed using SPSS for Win- dows (version 13.0; SPSS, Chicago, IL). Significance was accepted at P Ͻ 0.05. For the in vitro study, Student’s t test was used to evaluate the significance of the effect of sebacic acid and insulin on glucose transport. Results are expressed as means Ϯ SD of five different experi- ments, each performed in triplicate; SE is reported for percent values. Figure 1—Time course of plasma glucose (millimoles per liter), insulin (picomoles per liter), and RESULTS — Type 2 diabetic patients sebacic acid concentrations (micrograms per milliliter) in healthy volunteers (left panels) and in type 2 diabetic subjects (right panels). f, control study (0 g C10); E, 10 g C10 study; u,23gC10 (five women and five men) and control Ϯ subjects (four women and six men) were study. Data are means SE. As indicated by an arrow, the meal was ingested after 120 min of 6,6-D -glucose-primed constant infusion to assess basal glucose R . matched for sex distribution, age (52.1 Ϯ 2 a 6.98 vs. 47.2 Ϯ 6.03 years), and BMI Ϯ Ϯ 2 Ϫ (27.98 4.08 vs. 26.63 3.03 kg/m ). value of 39% in the 10 g C10 group and Glucose Ra AUC was decreased simi- Diabetic subjects were insulin resistant Ϫ71% in the 23 g C10 group (both P Ͻ larly in control and diabetic subjects after with a Matsuda index of 5.65 Ϯ 1.41 0.01) (Fig. 2). C10 supplement (by ϳ18% after 23 g compared with 15.29 Ϯ 2.63 in control The effect of C10 on plasma glucose C10, P Ͻ 0.05) (Fig. 2B), whereas glucose subjects (P Ͻ 0.01). and insulin time courses in type 2 diabetic clearance, which is an index of peripheral All subjects ingested a mixed meal subjects is depicted in Fig. 1. A reduction insulin sensitivity, tended to increase after (see RESEARCH DESIGN AND METHODS) contain- in the incremental glucose AUCs of 42 C10 supplementation, but reached statis- ing either 0 g (control) and an additional and 70% was observed in the 10 g C10 tical significance only in control subjects 10 g C10 or 23 g C10 as a substitute of (P ϭ 0.037) and 23 g C10 (P ϭ 0.045) after 23 g C10 (Fig. 2D). dietary fats. The time courses of plasma groups, respectively (Fig. 2). Similarly to As shown in Fig. 1, the concentration glucose and insulin in healthy control what was observed in control subjects, of plasma C10 in diabetic patients (Fig. subjects are shown in Fig. 1. The inges- the incremental AUCs of insulin were de- 1F) tended to be higher than that in con- tion of C10 together with the meal re- creased by 39% after 10 g C10 intake and trol subjects (Fig. 1E), without reaching duced to some extent the glycemic peak, by 64% after 23 g C10 intake, respectively statistical significance, however, and but the glucose incremental AUC was sig- (P Ͻ 0.05) (Fig. 2C). Insulin secretion peaked at the same times. Furthermore, nificantly reduced only after 23 g C10 rate was similar across the three studies in in diabetic subjects the two curves (10 vs. (Fig. 2). The insulin peak level was clearly both type 2 diabetic patients and healthy 23 g) overlapped up to 320 min after the reduced in both C10 groups, attaining a volunteers (Fig. 2). meal, and then the concentration of C10 care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 11, NOVEMBER 2010 2329 Oral sebacic acid in type 2 diabetes

A B 1000 6000

800 5000 * 4000 * 600 * 3000 AUC glucose AUC mol/ kg)mol/ ∆ ∆ ∆ ∆ 400 µ µ µ * µ ( 2000 200 (mmol/l x 300min) 1000 Positive

0 Rate ofAUC appearance 0 Control T2DM Control T2DM C D 35000 18

-1 * 30000 16 14 25000 12 20000 * 10 * 15000 8 AUC insulin AUC ∆ ∆ ∆ ∆ 6 10000 * * (pmol/l x 300min) 4 5000

ml/min.kg x (pmol/l) 2 0 0 Control T2DM insulin Clearance/AUC AUC Control T2DM E F 300 350 300 250

250 200 200

x 300 min) 150 AUC glucose 150 2 ∆ ∆ ∆ ∆ 100 AUC insulin 100 ∆ ∆ ∆ ∆ 50 50 (nmol/m Positive Positive 0 0 Control T2DM InsulinAUC secretion rate Control T2DM

Figure 2—Left panels: Incremental AUC of glucose (A) and insulin (C) concentrations after the meal and ratio between ⌬AUC of insulin and glucose

(E). Right panels: AUC of glucose Ra (B), AUC of glucose clearance rate normalized by AUC insulin concentrations (D), and insulin secretion rate (F). f, control study (0 g C10); , 10 g C10 study; p, 23 g C10 study. Data are means Ϯ SE. *Statistical significance (P Ͻ 0.05) between the study with and without C10. T2DM, type 2 diabetes. declined slower in the patients fed 23 g min [P ϭ 0.019]), corresponding to a per- betic subjects. This reduction occurs C10 compared with those fed 10 g C10. cent increase of 38.7 Ϯ 10.3% (mean Ϯ when C10 is either added to a mixed meal The in vitro experiments were con- SE) in presence of C10 vs. 11.4 Ϯ 5.4% in or substituted for lipids in an equivalent ducted by using a concentration of 0.2 its absence (P ϭ 0.026). This increase was caloric manner. The above effect is less mmol/l, on the basis of the average plasma associated with a 1.74 Ϯ 0.27-fold in- pronounced in healthy control subjects. concentration of C10 reached in humans crease of glucose transporter GLUT4 (Fig. In L6 cells, C10 significantly increases in- after C10 ingestion. In fact, a level of 40 3), in agreement with the slight increase sulin-mediated glucose uptake and ␮g/ml plasma C10, whose molecular in glucose clearance observed after C10 GLUT4 expression. weight is 202, corresponds to a concen- (Fig. 2D). There is general agreement that post- tration of 0.198 mmol/l. prandial hyperglycemia in type 2 diabetic The insulin-dependent glucose up- CONCLUSIONS — The major find- subjects mainly depends on a defect in the take from L6 cells increased more in the ing of the present study is that sebacic endogenous glucose output (EGO) sup- presence of C10 (1.83 Ϯ 0.40 vs. 1.37 Ϯ acid significantly reduces postmeal glu- pression. The defect in suppressing EGO ϭ 0.28 pmol/mg 10 min [P 0.01] vs. cose circulating levels as well as Ra in both after a meal has been estimated to account 1.73 Ϯ 0.19 vs. 1.57 Ϯ 0.14 pmol/mg 10 healthy and insulin-resistant type 2 dia- for 35–50% of the excess circulating glu-

2330 DIABETES CARE, VOLUME 33, NUMBER 11, NOVEMBER 2010 care.diabetesjournals.org Iaconelli and Associates

brane. This in vitro effect of C10 might partially explain the marked decrease in plasma glucose concentrations because it is well known that the skeletal muscle mass represents the most important con- sumer of glucose. We note that a true dose-response ef- fect of C10 on glucose disposal was not present in our investigation. In this re- gard, it is necessary to point out that, as previously demonstrated (25), the whole- body rate of C10 tissue uptake in male healthy volunteers was 180.9 Ϯ 4.5 ␮mol/min, which equals ϳ0.5 mg/kg body wt/min or 2.1 g/h in a subject weigh- ing 70 kg. Therefore, in the 5 h after C10 ingestion with the mixed meal, 10.5 g would be metabolized, thus explaining why the effects obtained with 10 or 23 g Figure 3—Left panel: Effect of sebacic acid on insulin-stimulated 2-deoxyglucose uptake by L6 were similar. cells. Serum-starved L6 myotubes were exposed to 0.2 mmol/l of sebacic acid overnight followed by In summary, although more detailed exposure to 10 nmol/l insulin for 20 min and then incubated with 2-[3H]deoxyglucose for 10 min studies are required to elucidate the at 37°C. Results are means Ϯ SE of five experiments performed in duplicate and are expressed as mechanism of action of sebacic acid in picomoles per milligram of protein during 10 min. P Ͻ 0.05 vs. insulin alone. Right panel: Western type 2 diabetes, this study provides rele- blot analysis of GLUT-4 expression in total protein extracts from L6 myoblasts treated with 0.2 vant clues about its effect in reducing glu- mmol/l sebacic acid overnight. Expression of tubulin was determined to ensure similar protein cose circulating levels when added to loading. y-axis represents arbitrary units. Results are means Ϯ SD. standard meals. cose in this state (20), the rest being at- sequent fall in others (22). Therefore, tributed to impaired postprandial insulin with all the precautions of incomplete in- Acknowledgments— The clinical study was stimulation of peripheral glucose uptake. formation, we can remark that sebacic sponsored by Nestec. The major contributor to the lack of EGO acid induces a significant reduction in the No other potential conflicts of interest rele- suppression and, thus, to the exaggerated glucose R , which can be ascribed to a vant to this article were reported. a A.I., D.G., and A.F. participated in the study postprandial hyperglycemia in diabetes reduced hepatic glucose output and/or to coordination, participant enrollment, medical seems to be an increase in gluconeogene- reduced glucose absorption. oversight of participants, and data collection. sis (21). A possible mechanism through which A.G. participated in the analysis of stable iso- Our study does not allow discrimi- sebacic acid might reduce EGO after a topes and reviewed/edited the manuscript. nating among systemic Ra of ingested glu- meal could be related to its mitochondrial C.C. participated in the in vitro studies. C.B. cose, postprandial endogenous glucose ␤-oxidation with production of succinic reviewed/edited the manuscript. K.M. partici- production, and peripheral tissue uptake. acid (10), which enters the mitochondrial pated in the conception and design of the study and reviewed/edited the manuscript. In fact, to measure the Ra of ingested glu- tricarboxylic acid cycle. Reduced in vivo cose, we should have used a double trac- mitochondrial function in skeletal muscle G.M. participated in the conception and de- er: one infused, as we did, and the other under resting conditions as well as during sign of the study and in the data analysis and interpretation and wrote the article. labeling the carbohydrate content in the postexercise recovery has been observed We thank Anna Caprodossi, Catholic Uni- meal. However, also in this case the Ra of in type 2 diabetic subjects (23). This im- versity, Department of Medicine, Rome, for meal-derived glucose would have been paired mitochondrial oxidative capacity her excellent technical assistance. influenced by a variety of factors, such as seems to be attributable to a reduced mi- the type of complex carbohydrates, the tochondrial tricarboxylic acid cycle flux rate of gastric emptying and orocecal tran- (24). C10 might overcome this defective References sit time, and the fraction extracted by the mitochondrial tricarboxylic acid function 1. Gastaldelli A, Baldi S, Pettiti M, Toschi E, liver, the latter being largely dependent in diabetic skeletal muscles by providing Camastra S, Natali A, Landau BR, Ferran- on the prevailing glucose and, to a lesser succinic acid and thus increasing whole- nini E. Influence of obesity and type 2 extent, insulin concentrations. In this re- body glucose clearance, skeletal muscle diabetes on gluconeogenesis and glucose gard, it is of note that when the meal is tissue being the major contributor to glu- output in humans: a quantitative study. composed of complex carbohydrates, the cose uptake. Diabetes 2000;49:1367–1373 glucose tracer can give a wide variability In this regard, we have observed that 2. Roden M, Bernroider E. Hepatic glucose of the estimates of initial splanchnic glu- in L6 cells, 0.2 mmol/l C10 significantly metabolism in humans: its role in health ϳ and disease. Best Pract Res Clin Endocri- cose uptake as well as of the pattern of increased glucose uptake by 30% com- nol Metab 2003;17:365–383 endogenous glucose production, which pared with insulin alone and increased 3. Wahren J, Ekberg K. Splanchnic regula- has been reported to be rapidly sup- GLUT4 protein expression 1.7-fold, tion of glucose production. Annu Rev pressed in some studies but to show an probably improving insulin signaling and Nutr 2007;27:329–345 initial paradoxical rise followed by a sub- the docking of GLUT4 to the cell mem- 4. Tappy L. Regulation of hepatic glucose care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 11, NOVEMBER 2010 2331 Oral sebacic acid in type 2 diabetes

production in healthy subjects and pa- tivity indices obtained from oral glucose dation are differentially impaired by the tients with non-insulin-dependent diabe- tolerance testing: comparison with the eu- IR1152 mutant receptor. J Biol Chem tes mellitus. Diabetes Metab 1995;21: glycemic insulin clamp. Diabetes Care 1997;272:7290–7297 233–240 1999;22:1462–1470 20. DeFronzo RA. Pathogenesis of type 2 5. Jenkins DJ, Ocana A, Jenkins AL, Wolever 13. Wolfe RR. Radioactive and Stable Isotope (non-insulin-dependent) diabetes melli- TM, Vuksan V, Katzman L, Hollands M, Tracers in Biomedicine: Principles and Prac- tus: a balanced overview. Diabetologia Greenberg G, Corey P, Patten R. Meta- tice of Kinetic Analysis. New York, Wiley- 1992;35:389–397 bolic advantages of spreading the nutrient Liss, 1992 21. Gastaldelli A, Toschi E, Pettiti M, Fras- load: effects of increased meal frequency 14. Greco AV, Mingrone G, Raguso C, cerra S, Quin˜ ones-Galvan A, Sironi AM, in non-insulin-dependent diabetes. Am J Tataranni A, Finotti E, Tacchino RM, Ca- Natali A, Ferrannini E. Effect of physio- Clin Nutr 1992;55:461–467 pristo E, De Gaetano A, Castagneto M. logical hyperinsulinemia on gluconeo- 6. Kolattukudy PE, Walton TJ, Kushwaha Metabolic effects and disposition of seba- genesis in nondiabetic subjects and in RP. Biosynthesis of the C18 family of cate, an alternate dicarboxylic fuel sub- type 2 diabetic patients. Diabetes 2001; cutin acids: omega-hydroxyoleic acid, strate. Ann Nutr Metab 1992;36:1–11 50:1807–1812 omega-hydroxy-9,10-epoxystearic acid, 15. Steele R. Influences of glucose loading 22. Basu R, Di Camillo B, Toffolo G, Basu A, 9,10,18-trihydroxystearic acid, and their and of injected insulin on hepatic glucose Shah P, Vella A, Rizza R, Cobelli C. Use of delta12-unsaturated analogs. Biochemis- output. Ann NY Acad Sci 1959;82:420– a novel triple-tracer approach to assess try 1973;12:4488–4498 430 postprandial glucose metabolism. Am J 7. Kolattukudy PE. Polyesters in higher 16. Van Cauter E, Mestrez F, Sturis J, Polon- Physiol Endocrinol Metab 2003;284: plants. Adv Biochem Eng Biotechnol sky KS. Estimation of insulin secretion E55–E69 2001;71:1–49 rates from C-peptide levels: comparison 23. Schrauwen-Hinderling VB, Kooi ME, 8. Kølvraa S, Gregersen N. In vitro studies of individual and standard kinetic param- Hesselink MK, Jeneson JA, Backes WH, on the oxidation of medium-chain dicar- eters for C-peptide clearance. Diabetes van Echteld CJ, van Engelshoven JM, boxylic acids in rat liver. Biochim Biophys 1992;41:368–377 Mensink M, Schrauwen P. Impaired in Acta 1986;876:515–525 17. Koivisto UM, Martinez-Valdez H, Bilan vivo mitochondrial function but similar 9. Nilsson A, Arey H, Pedersen JI, Chris- PJ, Burdett E, Ramlal T, Klip A. Differen- intramyocellular lipid content in patients tiansen EN. The effect of high-fat diets on tial regulation of the GLUT-1 and with type 2 diabetes mellitus and BMI- microsomal lauric acid hydroxylation in GLUT-4 glucose transport systems by matched control subjects. Diabetologia rat liver. Biochim Biophys Acta 1986;879: glucose and insulin in L6 muscle cells 2007;50:113–120 209–214 in culture. J Biol Chem 1991;266:2615– 24. Befroy DE, Petersen KF, Dufour S, Mason 10. Mingrone G, Castagneto M. Medium- 2621 GF, de Graaf RA, Rothman DL, Shulman chain, even-numbered dicarboxylic acids 18. Ueyama A, Yaworsky KL, Wang Q, Ebina GI. Impaired mitochondrial substrate ox- as novel energy substrates: an update. Y, Klip A. GLUT-4myc ectopic expression idation in muscle of insulin-resistant Nutr Rev 2006;64:449–456 in L6 myoblasts generates a GLUT-4-spe- offspring of type 2 diabetic patients. Dia- 11. Salinari S, Bertuzzi A, Gandolfi A, Greco cific pool conferring insulin sensitivity. betes 2007;56:1376–1381 AV, Scarfone A, Manco M, Mingrone G. Am J Physiol Endocrinol Metab 1999; 25. Mingrone G, Greco AV, Bertuzzi A, Dodecanedioic acid overcomes metabolic 277:E572–E578 Arcieri-Mastromattei E, Tacchino RM, inflexibility in type 2 diabetic subjects. 19. Caruso M, Miele C, Formisano P, Con- Marino F, Finotti E, Castagneto M. Tissue Am J Physiol Endocrinol Metab 2006; dorelli G, Bifulco G, Oliva A, Auricchio R, uptake and oxidation of disodium seba- 291:E1051–E1058 Riccardi G, Capaldo B, Beguinot F. In cate in man. J Parenter Enteral Nutr 1991; 12. Matsuda M, DeFronzo RA. Insulin sensi- skeletal muscle, glucose storage and oxi- 15:454–459

2332 DIABETES CARE, VOLUME 33, NUMBER 11, NOVEMBER 2010 care.diabetesjournals.org