Loss of Stearoyl-CoA Desaturase-1 Improves Insulin Sensitivity in Lean Mice but Worsens Diabetes in Leptin-Deficient Obese Mice

  1. Jessica B. Flowers12,
  2. Mary E. Rabaglia2,
  3. Kathryn L. Schueler2,
  4. Matthew T. Flowers12,
  5. Hong Lan2,
  6. Mark P. Keller2,
  7. James M. Ntambi12 and
  8. Alan D. Attie2
  1. 1Department of Nutritional Sciences, University of Wisconsin-Madison, Madison, Wisconsin
  2. 2Department of Biochemistry, University of Wisconsin-Madison, Madison, Wisconsin
  1. Address correspondence and reprint requests to Alan D. Attie, Department of Biochemistry, University of Wisconsin-Madison, 433 Babcock Dr., Madison, WI 537606. E-mail: attie{at}biochem.wisc.edu


The lipogenic gene stearoyl-CoA desaturase (SCD)1 appears to be a promising new target for obesity-related diabetes, as mice deficient in this enzyme are resistant to diet- and leptin deficiency–induced obesity. The BTBR mouse strain replicates many features of insulin resistance found in humans with excess visceral adiposity. Using the hyperinsulinemic-euglycemic clamp technique, we determined that insulin sensitivity was improved in heart, soleus muscle, adipose tissue, and liver of BTBR SCD1-deficient mice. We next determined whether SCD1 deficiency could prevent diabetes in leptin-deficient BTBR mice. Loss of SCD1 in leptinob/ob mice unexpectedly accelerated the progression to severe diabetes; 6-week fasting glucose increased ∼70%. In response to a glucose challenge, Scd1−/− leptinob/ob mice had insufficient insulin secretion, resulting in glucose intolerance. A morphologically distinct class of islets isolated from the Scd1−/− leptinob/ob mice had reduced insulin content and increased triglycerides, free fatty acids, esterified cholesterol, and free cholesterol and also a much higher content of saturated fatty acids. We believe the accumulation of lipid is due to an upregulation of lipoprotein lipase (20-fold) and Cd36 (167-fold) and downregulation of lipid oxidation genes in this class of islets. Therefore, although loss of Scd1 has beneficial effects on adiposity, this benefit may come at the expense of β-cells, resulting in an increased risk of diabetes.

Stearoyl-CoA desaturase (SCD) catalyzes the synthesis of monounsaturated fatty acids by introducing a cis double bond in the Δ9 position of saturated 16- and 18-carbon fatty acyl-CoA substrates. The products of SCD, palmitoleoyl-CoA and oleoyl-CoA, are the most abundant monounsaturated fatty acids of phospholipids, triglycerides, cholesterol esters, and wax esters (1). Various diseases, including cancer, obesity, diabetes, and atherosclerosis, are associated with an imbalance in the ratio of saturated to monounsaturated fatty acids (2). Two human and four mouse isoforms of SCD have been characterized. SCD1 is the primary isoform found in lipogenic tissues.

Loss of Scd1 is protective against obesity (35). The decreased adiposity in the Scd1−/− mice is due to reduced lipid synthesis and increased energy expenditure through enhanced fatty acid oxidation (3). Scd1−/− mice have an increase in insulin signaling and glucose uptake in muscle and brown adipose tissue (6,7).

Leptin-deficient BTBR leptinob/ob mice are a particularly useful animal model for studying obesity-related diabetes. Unlike the B6 leptinob/ob mice, which develop only moderate hyperglycemia, BTBR leptinob/ob mice become severely diabetic (810). The diabetic BTBR leptinob/ob mice have increased Scd1 expression in muscle and decreased expression in adipose tissue and liver, compared with nondiabetic B6 leptinob/ob mice (10).

Insulin resistance is often correlated with abdominal obesity. The BTBR mouse strain has excess abdominal obesity associated with insulin resistance in heart, soleus muscle, and adipose tissue (9,1113). In the present study, we measured the effect of SCD1 deficiency on insulin action in vivo during a 2-h hyperinsulinemic-euglycemic clamp in BTBR Scd1+/+ and Scd1−/− mice. We demonstrate that loss of SCD1 in BTBR lean mice significantly increases glucose disposal, primarily by increasing heart, soleus muscle, and adipose tissue insulin sensitivity.

To determine whether improved insulin sensitivity and reduced adiposity are sufficient to prevent obesity-induced diabetes, we generated BTBR leptinob/ob SCD1-deficient mice. We report that despite a reduction in body weight, loss of SCD1 unexpectedly accelerated the progression to severe diabetes in leptinob/ob mice.


BTBR T+ tf/J (BTBR) and C57BL/6J (B6) leptinob/+ animals were purchased from The Jackson Laboratory (Bar Harbor, ME) and bred at the University of Wisconsin Madison. BTBR Scd1−/− mice were created from SV129 Scd1−/− (14) mice using marker-assisted backcrossing (15) for at least six generations. BTBR leptinob/ob mice were independently created from B6 leptinob/+ mice using marker-assisted backcrossing for at least six generations (8). BTBR Scd1+/− were then bred to BTBR leptinob/+ mice. Resulting BTBR Scd1+/− leptinob/+ mice were intercrossed to generate BTBR leptinob/ob Scd1−/− mice. A similar strategy was used for generation of B6 leptinob/ob Scd1−/− mice. Mice were housed on a 12-h light-dark cycle, lights on at 0600 h. Mice had free access to a chow diet of roughly 6% fat (by mass; Purina 5008) and water except during fasts. At 6, 10, and 14 weeks of age mice were weighed and then fasted for 4 h. Following the 4-h fast ∼100 μl blood was collected from the retro-orbital sinus for measurements of plasma glucose, insulin, C-peptide, and triglyceride levels. Research protocols were approved by the University of Wisconsin Institutional Animal Care and Use Committee.

Hyperinsulinemic-euglycemic clamp.

Surgery, experimental procedures, and calculations have been previously described in detail (13). Scd1−/− mice required a high rate of glucose infusion. To minimize differences in infusion volume, a 20% solution was used for wild-type mice and a 40% solution for Scd1−/− mice.

Real-time quantitative PCR.

RNA was isolated from livers and islets using RNeasy mini columns with on-column DNase digestion (Qiagen). First-strand cDNA was synthesized from 1 μg total liver RNA or 0.2–0.5 μg islet RNA using Super Script III Reverse Transcriptase primed with a mixture of oligo-dT and random hexamers. Liver reactions were performed on an ABI GeneAmp 5700 Sequence Detection System (Applied Biosystems) and carried out in a 25-μl volume of 1× SYBR green PCR Core Reagents (Sigma-Aldrich) containing cDNA template from 10 ng of total RNA and 6 pmol primers. Islet reactions were performed on an ABI GeneAmp 7500 Sequence Detection System (Applied Biosystems) and carried out in a 20-μl volume of 1× ABI SYBR green PCR master mix containing cDNA template from 2 ng total RNA and 6 pmol primers. We determined the cycle at which the abundance of the accumulated PCR product crossed a specific threshold, the threshold cycle (CT), for each reaction. β-Actin was also used as a normalization control in islet measurements. Primer specificity was determined by observation of a single dissociation peak. A file with the sequences of the primers used in these studies is posted in the supplementary data section (available at http://dx.doi.org/10.2337/db06-1142).

Glucose tolerance test.

Following a 12-h fast (2000–0800 h), mice were given an intraperitoneal injection of a 30% glucose solution, 2 mg/g body wt. Blood samples ∼30 μl were collected from the tail at t = 0, 15, 30, 60, and 120 min for the determination of plasma glucose and insulin.

Islet isolation.

Following a 4-h fast, intact pancreatic islets were isolated from mice using a collagenase digestion procedure (16). A Ficoll gradient was used to partially purify islets from digested pancreata. For further purification, islets were hand-picked using a stereomicroscope. Islets not used for secretion experiments were rinsed twice with 1× PBS, homogenized in RLT buffer, and stored at −80°C until further analysis.

Islet insulin secretion.

Three islets of equivalent size were placed in 12 × 75 mm glass tubes, where the bottom of the tube was formed by a 62-μm mesh. The 12 × 75 mm tubes were transferred to 16 × 100 mm tubes containing 1 ml Kreb's Ringer buffer with 1.7 mmol/l glucose and 0.5% BSA and preincubated at 37°C for 45 min. Following the preincubation, the 12 × 75 mm tubes were transferred to fresh 16 × 100 mm tubes containing 1 ml Kreb's Ringer buffer supplemented with 1.7 or 16.7 mmol/l glucose. Following a 45-min incubation period at 37°C, the 12 × 75 mm tubes were transferred to a fresh tube containing 1 ml HCl ethanol water (1:50:14) to extract cellular insulin from the islets. The media left in the 16 × 100 mm tubes was collected and frozen for insulin determination by enzyme-linked immunosorgent assay.

Islet lipid composition.

Following isolation described above, lipids were extracted from 20–100 μg islet protein (∼100 islets) and separated by thin- layer chromatography (17). The triglyceride and free fatty acid (FFA) bands were scraped into tubes containing pentadecanoic acid (internal standard). Fatty acids were then trans-methylated and analyzed by gas liquid chromatography (18). The cholesterol and cholesterol ester bands were scraped, extracted, diluted with 90 μl of 10% Triton-X100 in isopropyl alcohol, and measured with the Wako Cholesterol E kit (19).

Plasma lipid analysis.

Plasma lipoproteins were fractionated on a Superose 6HR 10/30 FPLC column (Amersham Bioscience) as described (19). The equivalent of 100 μl plasma was injected onto the column, and 500-μl fractions were collected and assayed for total cholesterol and triglyceride.


Data represent means ± SE. Comparisons among three groups were made using one-way ANOVA, and when significance was observed, post hoc Tukey tests were performed using GraphPad Prism version 4.02 for Windows (GraphPad Software, San Diego, CA). Comparisons of two groups were done using unpaired Student's t test. If an F test to compare variances revealed a P value <0.01, data were log transformed before analysis, designated PLOG. Data analyzed by Mann-Whitney rank sum test are designated PGAUS to indicate that the P values are Gaussian approximations.


Impact of SCD1 deficiency on in vivo glucose disposal and on the partitioning of glucose fluxes.

Lean BTBR Scd1−/− mice exhibit significantly lower fasting/basal plasma glucose and insulin (Fig. 1A and B), suggesting enhanced insulin sensitivity. In hyperinsulinemic-euglycemic clamp experiments, the rate of glucose infusion needed to maintain euglycemia at ∼100 mg/dl was threefold higher in Scd1−/− than wild-type mice (Fig. 1C). Scd1−/− mice had a twofold higher whole-body glucose disposal rate than wild-type mice (Fig. 1C). Two processes contribute to glucose disposal: glycogen synthesis and glycolysis. Glycolysis, which accounted for the majority of glucose disposal, was increased twofold in Scd1−/− mice (Fig. 1C). Glycogen synthesis was also higher in Scd1−/− mice than in wild-type mice (Fig. 1C). Thus, loss of Scd1 in lean mice significantly increased glucose disposal by increasing glycolysis and glycogen synthesis.

To determine which tissues are responsible for the increased insulin-stimulated glucose utilization in Scd1−/− lean mice, we measured the rate of insulin-stimulated 2-deoxyglucose uptake in individual tissues during the hyperinsulinemic clamp. Whereas white adipose tissue glucose uptake tended to be higher in Scd1−/− mice (P = 0.10, Fig. 1D), quadriceps muscle failed to show differences in glucose uptake; however, 2-deoxyglucose uptake of Scd1−/− mice was markedly increased in heart and soleus muscle (Fig. 1D).

Hepatic gene expression and hepatic glucose output.

Decreased hepatic lipogenic gene expression (3) and lipogenesis (3,5,20,21) have previously been noted in Scd1−/− animals. In liver, substrates may be diverted from lipogenesis to gluconeogenesis. However, decreased hepatic lipogenic gene expression (Fig. 1E) was associated neither with an increase in gluconeogenic gene expression (Fig. 1E) nor with an increase in hepatic glucose output (HGO) (Fig. 1F). Under fasting conditions, wild-type and Scd1−/− mice have similar rates of HGO (Fig. 1F). During insulin infusion, wild-type mice showed a normal 74% inhibition of HGO, whereas Scd1−/− mice showed enhanced liver insulin sensitivity, complete inhibition of HGO, and significantly lower expression of G6Pase (Fig. 1E and F). Insulin sensitivity was similar in Scd1+/− and Scd1−/− mice (Fig. 1G). Partial inhibition of SCD1 may also be effective in treating abdominal obesity; Scd1+/− and Scd1−/− mice both had a 50% decrease in epididymal white adipose tissue (Fig. 1H).

Impact of SCD1 deficiency on leptin-deficient mice.

To determine whether reduced adiposity and increased insulin sensitivity observed in lean mice would translate to improved diabetes symptoms in obese mice, we generated Scd1−/− leptinob/ob mice. Despite a 20% reduction in body weight (Table 1), insulin levels in Scd1−/− leptinob/ob mice were 38% lower than in Scd1+/+ leptinob/ob mice at 6 weeks of age (Table 1). The lower insulin levels were accompanied by increased glucose and fasting triglyceride levels (Table 1). In contrast, Scd1 deficiency had no effect on plasma triglyceride levels in lean mice.

Between 6 and 10 weeks of age, BTBR mice become progressively more diabetic (810). By 10 weeks, fasting glucose exceeds 300 mg/dl. Due to the severity of diabetes in older mice, studies with BTBR mice were carried out in 6- to 7-week-old mice. Insulin levels were lower and glucose levels higher in Scd1−/− leptinob/ob mice than in Scd1+/+ leptinob/ob mice (Table 1, Fig. 2A). B6 mice are less prone to diabetes than BTBR mice (810). Consequently, genetic factors present in the B6 strain delayed onset and protected some Scd1−/− leptinob/ob mice from developing diabetes. Nonetheless, 70% of B6 Scd1−/− leptinob/ob mice had fasting glucose levels >300 mg/dl and insulin levels <15 ng/ml at 14 weeks, compared with 17% of Scd1+/+ leptinob/ob mice (Fig. 2B). Thus, despite improving insulin sensitivity in lean mice, loss of Scd1 worsens diabetes in leptinob/ob mice.

In vivo insulin secretion and glucose clearance.

To determine whether the reduced insulin seen in Scd1−/− leptinob/ob mice is due to a defect in insulin secretion, we monitored insulin during a glucose tolerance test. BTBR Scd1−/− leptinob/ob mice were less tolerant to a glucose challenge than Scd1+/+ leptinob/ob mice (Fig. 2C and D). Insulin secretion during the glucose tolerance test was significantly reduced in Scd1−/− leptinob/ob mice (Fig. 2D).

Insulin secretion from isolated islets.

We compared the responsiveness of islets from 6- to 7-week-old BTBR Scd1+/+ leptinob/ob and Scd1−/− leptinob/ob mice to low (1.7 mmol/l) and high (16.7 mmol/l) glucose. Following islet isolation, we observed two distinct classes of islets in Scd1−/− leptinob/ob mice, designated “type A” and “type B” (Fig. 3A). We compared the insulin secretion response of type A islets from Scd1+/+ and Scd1−/− leptinob/ob mice, as all islets isolated from Scd1+/+ were of this type.

Total insulin secretion from type A islets was similar in Scd1+/+ and Scd1−/− mice (Fig. 3C). When expressed as a fraction of islet insulin content, insulin secretion tended to be higher in response to 16.7 mmol/l glucose in Scd1−/− mice (P = 0.10, Fig. 3D) because islet insulin content tended to be lower (P = 0.15, Fig. 3E). Islet insulin content was further reduced in type B islets from Scd1−/− mice (Fig. 3E). We were unable to detect measurable levels of insulin released from type B islets after incubation in 16.7 mmol/l glucose. The high prevalence of type B islets, ∼50% of the total islets isolated from Scd1−/− mice (Fig. 3F), is consistent with the 33% reduction in whole pancreas insulin content of Scd1−/− mice (Fig. 3G). Therefore, although there was not a defect in islet insulin secretion per se, the >50% reduction in insulin secretion during the glucose tolerance test is proportional to the relative number of type B islets and the reduced pancreatic insulin content of the Scd1−/− leptinob/ob mice.

Type B islets were also observed in 14-week-old Scd1−/− leptinob/ob B6 mice, and pancreatic insulin content was significantly lower in B6 Scd1−/− compared with Scd1+/+ leptinob/ob mice (Fig. 3H). Thus, the observed pancreatic defect is not specific to the diabetes-susceptible BTBR mice but is also present in B6 Scd1−/− mice, which are less susceptible to diabetes.

Plasma lipid composition of Scd1−/−leptinob/ob mice.

Chronic exposure of β-cells to high levels of fatty acids, particularly saturated fatty acids, is associated with impaired glucose-stimulated insulin secretion and eventually β-cell death (2228). We hypothesized that β-cells from Scd1−/− leptinob/ob mice are dysfunctional due to chronic exposure to a higher ratio of saturated fatty acid (SFA) to monounsaturated fatty acid (MUFA). To determine the circulating levels of SFA and MUFA in Scd1+/+ leptinob/ob and Scd1−/− leptinob/ob mice, we measured the FFA and triglyceride concentrations of the substrates and products of SCD1. The concentrations of palmitoleate (16:1) and oleate (18:1 n-9) were significantly lower in Scd1−/− leptinob/ob mice (Fig. 4A), whereas the concentration of stearate (18:0) was significantly elevated (Fig. 4A). Similar trends were observed for plasma triglyceride, although only oleate was significantly reduced in plasma from Scd1-deficient leptinob/ob mice (Fig. 4C). Overall, the plasma percentage of SFA was significantly elevated in both FFA and triglyceride from Scd1−/− leptinob/ob mice (Fig. 4B and D).

Exposure of islets to LDL cholesterol can elicit β-cell necrosis (29). Leptin-deficiency significantly elevated total plasma cholesterol, but to a lesser extent in Scd1−/− leptinob/ob mice (Fig. 4E). Whereas the cholesterol elevation in Scd1−/− leptinob/ob mice was primarily due to increased VLDL and LDL cholesterol, Scd1+/+ leptinob/ob mice had elevated HDL in addition to VLDL and LDL cholesterol (Fig. 4F). Given that both LDL and HDL were reduced in Scd1−/− leptinob/ob mice compared with Scd1+/+ leptinob/ob mice, it is not clear if this would promote dysfunction or protection of the β-cells in Scd1−/− leptinob/ob mice.

Islet lipid composition.

We noted that the type B islets were more buoyant than type A islets. Whereas the FFA oleate concentration was not significantly different between the two islet types, palmitoleate was increased 3.5-fold, and palmitate and stearate were increased threefold in type B islets (Fig. 5A). The most abundant triglyceride fatty acid, palmitate, was increased 4.5-fold in type B relative to type A islets. The other triglyceride fatty acids were similarly increased, but to a lesser extent (Fig. 5B). To determine whether the reduced plasma cholesterol in Scd1-deficient leptinob/ob mice was associated with a change in islet cholesterol content, we measured the level of free and esterified cholesterol extracted from islets. Compared with type A islets from Scd1+/+ leptinob/ob mice, total cholesterol tended to be increased in type A islets from Scd1−/− leptinob/ob mice (P = 0.09, Fig. 5C) and was significantly elevated (almost sixfold) in type B islets from Scd1−/− leptinob/ob mice. Whereas type A islets from Scd1−/− leptinob/ob mice only had increased cholesterol ester compared with Scd1+/+ leptinob/ob type A islets, type B islets had significantly elevated free and esterified cholesterol (Fig. 5C). Hence, there is a severe impairment in lipid homeostasis in islets from Scd1-deficient leptinob/ob mice.

Islet expression of lipid metabolism genes.

Despite the extreme cellular stress caused by cholesterol accumulation (30), we were able to isolate high-quality RNA from both type A and type B islets from Scd1−/− leptinob/ob mice (Fig. 6A) and examined islet mRNA expression levels of genes involved in lipid synthesis, uptake, and oxidation (Table 2). Genes involved in lipid metabolism were downregulated in type B islets.

β-Cells express fatty acid transporters and Gpr40, a presumed fatty acid receptor (3133), and they also express lipoprotein lipase (Lpl) (34,35). Acutely, increased fatty acid availability augments glucose-stimulated insulin secretion; however, chronic exposure of β-cells to high levels of fatty acids impairs their function (3638). Overexpression of Gpr40 in β-cells results in impaired insulin secretion and diabetes (31). Similarly, INS1 β-cells that overexpress Lpl have reduced insulin secretion (34).

Lpl expression was increased in both type A (1.7-fold, P < 0.05) and type B (20-fold, P < 0.001) islets from Scd1−/− leptinob/ob mice, suggesting that dysregulation of Lpl expression is one of the earliest detectable differences in islets from Scd1−/− leptinob/ob mice (Table 2 and Fig. 6D). In contrast, Gpr40 expression was downregulated in both types of islets from Scd1−/− leptinob/ob mice, 41% in type A and 98% in type B (Table 2). Conversely, the expression of another fatty acid transporter, Cd36, was dramatically increased in type B islets (167-fold, P < 0.001; Table 2 and Fig. 6B). Failure to upregulate lipid oxidation genes may also contribute to the lipid accumulation observed in type B islets. The lipid oxidation genes Cpt1b, Hadhsc, Ppara, Ppargc1a, and Ucp2 were all downregulated in type B islets (Table 2).

Several changes in gene expression are consistent with the cholesterol overload phenotype observed in islets from Scd1−/− leptinob/ob mice (Figs. 5C and 6BF). Consistent with feedback regulation and reduced SREBP2 maturation (39), expression of LDL receptor (Ldlr) and hydroxymethylglutaryl-CoA synthase (Hmgcs1) were downregulated in type B islets (Fig. 6E and F). Additionally, LXR (liver X receptor) may be activated in type B islets by cholesterol-derived oxysterols; two LXR targets, phospholipid transfer protein (Pltp) and Lpl, were upregulated in type B islets (Fig. 6C and D). Cd36 expression is also enhanced by cholesterol uptake (40) and was increased in type B islets (Fig. 6B).

Islet expression of functional and death-associated genes.

Hyperglycemia and islet fatty acid and cholesterol accumulation have been previously associated with impaired β-cell function and cell death (22,24,25,29,30,4143). Consistent with altered β-cell function, GLUT2 (Slc2a2), insulin (Ins2), insulin promoter factor 1 (Ipf1/Pdx1), glucokinase (Gck), and insulin receptor substrate 2 (Irs2) were all dramatically (>90%; Table 2) downregulated in type B islets from Scd1−/− leptinob/ob mice. Cell disruption may not be restricted to β-cells; glucagon (Gcg) was also 95% lower in type B islets (Table 2). Insulin and Glut2 expression also tended to be lower in type A islets isolated from Scd1−/− leptinob/ob mice (41 and 47%, respectively, compared with type A islets from Scd1+/+ leptinob/ob mice; P = 0.053 and P = 0.131, Table 2), suggesting that these changes might be early indicators of β-cell dysfunction in obese Scd1-deficient mice.

To assess the possible mechanisms of cell death induced in the dysfunctional islets, we assessed the expression of genes that have been related to either protection from (Bcl2) or promotion of cell death [Casp1, Tnfa, Casp3, Bad, Cdkn1a(P21), Bax, Ddit3(Chop10), and Hspa5(Bip)]. The antiapoptotic gene Bcl2 was downregulated 50% in type B islets isolated from Scd1−/− leptinob/ob mice (Table 2). Paradoxically, several proapoptotic genes, Casp3, Bad, Cdkn1a, Bax, Ddit3, and Hspa5, were downregulated 67–93% in type B islets (Table 2). Two proapoptotic genes, Casp1 and Tnfa, were upregulated 35- and 6-fold, respectively, in type B islets (Table 2).


Scd1 deficiency results in reduced body weight in models of diet and leptin deficiency–induced obesity (35). Increased expression of Scd1 in muscle from obese humans and rats is associated with insulin resistance (44,45). Several studies, including this one, have also demonstrated that Scd1 deficiency can improve insulin sensitivity, making it a promising drug target for the treatment of insulin resistance (6,7). Whereas short-term treatment with an antisense inhibitor of Scd1 is sufficient to improve hepatic insulin sensitivity (46), longer treatment additionally prevents diet-induced obesity and is associated with lowered plasma glucose and insulin (47). For the first time, using the hyperinsulinemic-euglycemic clamp technique, we show that insulin-stimulated whole-body, heart, and soleus muscle glucose uptake are increased in Scd1−/− mice.

In lipodystrophic mice, loss of Scd1 promotes diabetes, despite normalizing hepatic lipid levels (20). Cohen et al. (5) have shown that loss of Scd1 in leptin-deficient obese mice is sufficient to normalize hepatic lipid storage and significantly reduce adiposity. Here, we demonstrate that despite reducing body weight, loss of Scd1 in leptin-deficient obese mice causes hypoinsulinemia and hyperglycemia.

We found that Scd1−/− leptinob/ob mice secrete less insulin than Scd1+/+ leptinob/ob mice. We discovered that whereas normal-looking islets isolated from Scd1−/− mice did not show an insulin secretion defect per se, a distinct class of islets isolated from the Scd1−/− leptinob/ob mice had reduced insulin content and increased triglyceride, FFAs, esterified cholesterol, and free cholesterol.

Diminished β-cell function and loss of β-cell mass are characteristic features of the progression from insulin resistance to type 2 diabetes. Pancreatic β-cells are particularly sensitive to SFA; chronic exposure of β-cells to fatty acids in vitro alters secretory function and induces apoptosis (2227). The degree of saturation of the fatty acids is important; SFAs cause marked apoptosis that can be prevented by unsaturated fatty acids (24). Busch et al. (48) have identified subpools of cultured β-cells that are resistant to the cytotoxic effects of the SFA palmitate. Of particular relevance to this study, they found that the palmitate-resistant cells desaturate palmitate by upregulating Scd1. An inhibitor of Scd1 dose-dependently reversed the resistance of palmitate-resistant cells to the cytotoxic effects of palmitate. We determined that the fatty acid composition of plasma and islets from Scd1−/− leptinob/ob mice was significantly more saturated.

Uptake of lipoprotein particles may provide an alternate route for β-cell lipotoxicity. Whereas CD36 primarily acts as a high-affinity fatty acid transporter in peripheral tissues (4952), it is also a receptor for oxidized LDL in macrophages (53,54). LDL is specifically taken up and degraded by β-cells but not other islet cell types (55). LDL induces necrosis in primary rat β-cells in vitro (29). Islet LDL uptake mediated by Cd36 has not previously been examined. However, Cd36 is expressed in human islets, as well as MIN6 β-cells (56). Therefore, the >150-fold upregulation in expression of CD36 in Scd1−/− leptinob/ob islets may mediate both the cholesterol and lipid accumulation that were observed in islets from these mice and may contribute to their death (57).

Consistent with cholesterol overload in these cells, we detected an upregulation in expression of Cd36, Pltp, and Lpl and a downregulation in expression of Ldlr and Hmgcs1. The type B islets also showed a marked reduction in the expression of Ppara and its target genes, which are involved in detoxification of lipids through fatty acid oxidation. This would tend to exacerbate the lipotoxicity evoked by the high lipid uptake and increase in SFAs in the β-cells.

We also detected a downregulation of Bcl2, an antiapoptotic gene, and upregulation of Casp1 and Tnfa, which mediate cell death. Macrophage infiltration of islets and local production of cytokines, including tumor necrosis factor, contribute to destruction of β-cells in type 1 diabetes. Interestingly, lipid-laden islets from ZDF rats, or islets chronically cultured in fatty acids, are more susceptible to cytokine-induced toxicity (58). We observed a high proportion of TUNEL (transferase-mediated dUTP nick-end labeling)-positive islets in the Scd1-deficient obese mice but did not carry out these studies in a sufficient number of animals to report quantitative data. These studies, and the results presented here, suggest that this mechanism may also contribute to β-cell destruction in type 2 diabetes.

Consistent with previously published reports, we show improved insulin sensitivity in lean Scd1-deficient mice. However, in genetically obese mice, hyperlipidemia induced by leptin deficiency leads to a dysregulation of lipid homeostasis in islets from Scd1-deficient mice. The lipid accumulation may be mediated by a >150-fold upregulation in expression of the cholesterol and fatty acid transporter Cd36. Given that the target population for Scd1 inhibitors is likely to be both obese and dyslipidemic, the beneficial effects of Scd1 inhibition on adiposity may come at the expense of β-cells, resulting in an increased risk of diabetes.

FIG. 1.

Effect of SCD1 deficiency on whole-body glucose metabolism and hepatic gene expression. A: Plasma glucose. B: Plasma insulin. C: Glucose flux. D: Tissue glucose uptake. E: Hepatic gene expression in fasted mice. F: Hepatic glucose output. G: Glucose disposal. H: Fat pad mass. *P < 0.05 vs. Scd1+/+, +P < 0.05 vs. basal, t test. G and H, *P < 0.05 by one-way ANOVA with Tukey post-test. Data represent mean ± SE of 8 mice (basal) and 4–6 mice (clamp).

FIG. 2.

Basal and postchallenge glucose and insulin values. A: 4-h fasting plasma glucose and insulin values from 6-week-old BTBR leptinob/ob male and female mice. B: 4-h fasting plasma glucose and insulin values from 14-week-old B6 leptinob/ob male and female mice. Dotted lines represent the average insulin and glucose values for Scd1+/+ leptinob/ob mice in each comparison. Individual values shown for Scd1+/+ leptinob/ob mice (gray) and Scd1−/− leptinob/ob mice (black). C and D: Glucose and insulin during an intraperitoneal glucose tolerance experiment. n = 6–8 with two males in each group. Bar graphs represent the area under the curve; for glucose P = 0.01, for insulin P = 0.03, t test. Data are represented as means ± SE.

FIG. 3.

Islet and pancreas characterization of leptin-deficient mice. A: Representative micrographs of all islets collected from a BTBR Scd1−/− leptinob/ob mouse (mixed population), as well as hand-picked type A and type B islets from the same animal (middle and right panels, respectively). B and C: Insulin secretion in response to low (1.7 mmol/l) and high (16.7 mmol/l) glucose from type A islets. Bars represent the mean ± SE of at least three separate experiments in triplicate. D: Pancreatic islet insulin content. E: Islet types in wild-type and Scd1−/− mice. F: Whole pancreas insulin content of 6-week-old BTBR mice. n = 8–9 with 2–3 males per group. P = 0.04. G: Whole pancreas insulin content of 14-week-old B6 mice. n = 9–10 with 2–3 males per group. PLOG < 0.0001, t test.

FIG. 4.

Effect of Scd1 deficiency on plasma fatty acid and cholesterol composition in 6-week-old leptinob/ob mice. Plasmas from fasting Scd1+/+and Scd1−/− (n = 5–8) male mice. A and C: ▪, Scd1+/+ leptinob/ob mice; □, Scd1−/− leptinob/ob mice. B and D: ▒, unsaturated fatty acids. E: Different letters designate a statistically significant difference by one-way ANOVA with Tukey post-test; P < 0.05. Data are represented as means ± SE. *P < 0.05 by t test. TG, triglyceride.

FIG. 5.

Effect of Scd1 deficiency on islet fatty acid and cholesterol composition in 6-week-old leptinob/ob mice. Lipids were extracted from 20–100 μg islet protein (∼100 islets). Data represent means ± SE with type A Scd1+/+ (▪, n = 4), Scd1−/− islets (□, n = 4) vs. type B Scd1−/− islets (▒, n = 6). *P < 0.05 by t test. Different letters designate a significant difference by ANOVA with Tukey post-test, P < 0.05.

FIG. 6.

Islet expression of genes related to the cholesterol overload phenotype. A: RNA isolated from type A and type B islets assessed by Agilent 2100 bioanalyzer. (RNA from A and B islets from the same animal are indicated.) mRNA levels were normalized to β-actin and are expressed relative to Scd1+/+ type A islets. B: CD36 antigen (CD36). C: phospholipid transfer protein (Pltp). D: lipoprotein lipase (Lpl). E: LDL receptor (Ldlr). F: 3-hydroxy-3-methylglutaryl-Coenzyme A synthase 1 (Hmgcs1). Letters designate significant differences by ANOVA with Tukey post-test. **P = 0.008 by t test. To minimize variance differences, β-actin–normalized data (∼log2 transformation of relative data) were used for statistical analyses.


Physiological measurements in 6-week-old leptin-deficient mice by phenotype


Islet levels of genes involved in lipid metabolism, islet function, and cell death


This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK58037 and DK66369 and National Heart, Lung, and Blood Institute Grant HL56593. J.F. was supported by Nutritional Sciences Departmental Predoctoral Training Grant NIH DK-07665-11.

We thank Susanne Clee, Agnieszka Dobrzyn, Angie Oler, Donnie Stapleton, and Anna Szabo for technical assistance.


  • Published online ahead of print at http://diabetes.diabetesjournals.org on 16 March 2007. DOI: 10.2337/db06-1142.

    Additional information for this article can be found in an online appendix at http://dx.doi.org/10.2337/db06-1142.

    M.P.K., J.M.N., and A.D.A. hold stock in Xenon Pharmaceuticals.

    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 U.S.C. Section 1734 solely to indicate this fact.

    • Accepted February 9, 2007.
    • Received August 15, 2006.


| Table of Contents

This Article

  1. Diabetes vol. 56 no. 5 1228-1239
  1. Online-Only Appendix
  2. All Versions of this Article:
    1. db06-1142v1
    2. 56/5/1228 most recent