12-Lipoxygenase-knockout mice are resistant to inflammatory effects of obesity induced by western diet

Abstract

Inflammation is a key pathological process in the progression of atherosclerosis and type 2 diabetes. 12/15-lipoxygenase (12-LO), an enzyme involved in fatty acid metabolism, may contribute to inflammatory damage triggered by stressors such as obesity and insulin resistance. We hypothesized that mice lacking 12-LO are protected against inflammatory-mediated damage associated with a “western” diet. To test this hypothesis, age-matched male 12-LO knockout (12-LOKO) and wild-type C57BL/6 (B6) mice were fed either a standard chow or western diet and assessed for several inflammatory markers. Western-fed B6 mice showed expected reductions in glucose and insulin tolerance compared with chow-fed mice. In contrast, western-fed 12-LOKO mice maintained glucose and insulin tolerance similar to chow-fed mice. Circulating proinflammatory cytokines, tumor necrosis factor-α and interleukin-6, were increased in western B6 mice but not 12-LOKO mice, whereas the reported protective adipokine, adiponectin, was decreased only in western B6 mice. 12-LO activity was significantly elevated by western diet in islets from B6 mice. Islets from 12-LOKO mice did not show western-diet-induced islet hyperplasia or increases in caspase-3 apoptotic staining observed in western-fed B6 mice. Islets from 12-LOKO mice were also protected from reduced glucose-stimulated insulin secretion observed in islets from western-fed B6 mice. In visceral fat, macrophage numbers and monocyte chemoattractant protein-1 expression were elevated in western B6 mice but not 12-LOKO mice. These data suggest that 12-LO activation plays a role in western-diet-induced damage in visceral fat and islets. Inhibiting 12-LO may provide a new therapeutic approach to prevent inflammation-mediated metabolic consequences of excess fat intake.

visceral adiposity is a contributing factor to both type 2 diabetes (T2D) and atherosclerosis. “Western” diets that are high in fat can lead to obesity, increased visceral adiposity, and hyperlipidemia. Chronic exposure to high lipid levels triggers an inflammatory response that can damage the endocrine pancreas, fat, and other organs. The inflammatory mediators include cytokines, reactive oxygen species, and lipid factors that promote insulin resistance, pancreatic islet dysfunction, and the formation of atherosclerotic plaques (7, 10, 11, 14, 27, 30, 35, 37, 49, 50, 52).

One potential player in high-fat-diet-induced inflammation is 12/15-lipoxygenase (12-LO). 12-LO is primarily involved in metabolizing arachidonic acid by inserting molecular oxygen at the 12-carbon atom to form the lipid 12-hydroperoxyeicosatetraenoic acid, which is subsequently converted to the more stable form 12-hydroxyeicosatetraenoic acid (12-HETE), and metabolizing linoleic acid to form 13-hydroxyoctadeca dienoic acid (13-HODE) (4). Increased 12-LO expression or 12-LO activity increases the production of 12-HETE, and these HETE products can activate c-Jun-NH₂ kinase and cyclooxygenase-2 signaling pathways related to increased localized oxidative stress (2, 21, 59). 12-LO expression and/or activity are upregulated by stressors, such as hyperglycemia (33, 39, 40, 60) or cytokine-mediated damage (13).

12-LO may play a direct role in inflammatory damage caused by a high-fat diet and the pathology of β-cell dysfunction in T2D. 12-LO is expressed in rodent and human islets (5, 46), and studies have linked hyperglycemia (33, 40, 60) and hyperlipidemia (13, 47) with increased 12-LO expression in islets. 12-LO is one of the key genes upregulated in hyperglycemic mice in the partial-pancreatectomy model of T2D (33), and Diabetic Zucker Fatty rats also demonstrate elevated 12-LO associated with deficits in insulin secretion (60). Furthermore, increased 12-LO expression and 12-LO activity (12-HETE production) in response to interleukin stimulation has been demonstrated in rodent islets and cell lines (5, 13), resulting in proinflammatory pathway activation by 12-LO products (20, 64). These findings support the hypothesis that 12-LO activation plays a role in inflammatory effects in β-cells, and could thus be a contributing factor to β-cell dysfunction observed in T2D (18).

12-LO also plays an active role in atherosclerosis via inflammatory actions on visceral adipose tissue. Visceral fat, but not subcutaneous fat, is associated with increased cardiovascular risk (32, 54, 55), possibly through increased leukocyte infiltration and adipose macrophage activity (62, 66, 67). 12-LO products have been shown to enhance the interaction of monocytes with vascular endothelium, which is involved in oxidative modification of lipoproteins and membrane lipids and initiating or accelerating atherogenesis (16, 21, 43). Furthermore, we have recently shown that 12-LO products upregulate expression of monocyte chemoattractant protein (MCP-1, also called CCL2) in mouse peritoneal macrophages and that 12/15-LO null mice have reduced MCP-1 expression (65). Increased MCP-1 levels have been observed in adipose tissue of obese rodents and humans, and MCP-1 appears to play an important role in mediating macrophage infiltration in adipose tissue and atherosclerotic plaques (24, 65), although recent findings suggest that MCP-1 may not be the only factor to macrophage infiltration (31).

These findings collectively suggest that 12-LO may link visceral adiposity, inflammation, and atherosclerosis. In this study, we investigated the effects of high-fat-diet-induced inflammation on 12-LO activation and subsequent metabolic effects by comparing 12-LOKO mice (6, 56) on the C57BL/6 (B6) background and wild-type B6 mice. Weight- and age-matched male mice at age 6–8 wk were placed on an 8-, 12-, or 24-wk diet of either standard chow or a high-fat western diet. We monitored body weight and blood glucose throughout the diet and measured glucose and insulin tolerance as well as serum adipokine levels at the end of the trial. We also analyzed glucose stimulation, apoptotic rates, 12-LO expression, and histology of pancreatic islets, as well as macrophage infiltration and MCP-1 levels in visceral adipose tissue. We hypothesized that 12-LO activation may play a role in promoting inflammation and that reducing 12-LO activity may prevent toxic effects of obesity induced by high-fat intake.

MATERIALS AND METHODS

Animals and treatments.

Male B6 mice were purchased from Jackson Laboratory (Bar Harbor, ME), and 12/15-lipoxygenase-null mice (12-LOKO) were originally a generous gift of Dr. Colin Funk (6, 56). The 12-LO mice have been backcrossed to the B6 background for at least six generations before inbreeding for homozygosity in the experimental mice. Microsatellite testing has confirmed >96% homology between the 12-LOKO and the C57BL/6J. Mice were placed on a chow or a high-fat “western-type” diet beginning at 6–8 wk of age. The chow and western diet foods were purchased from Harlan Teklad (Madison, WI); the western diet consisted of 42% of calories from fat, 15.3% of calories from protein, and 42.7% of calories from carbohydrate, primarily sucrose (TD#88137). All mice were placed on the diet for 8, 12, or 24 wk, during which time blood glucose and body weight were measured weekly before 10:00 A.M. Blood glucose was measured using Accu-Chek Advantage Glucose Monitors (Roche, Indianapolis, IN) with tail vein blood samples. Body weight was measured by placing the mice directly on a scale, and the same scale was used for each set of measurements. Mice were housed in a pathogen-free facility at the University of Virginia (UVA). All experiments were performed in accordance with an experimental protocol for animal study approved by the UVA Institutional Animal Care and Use Committee.

Glucose tolerance test and insulin tolerance test.

For the glucose tolerance test (GTT), mice were injected intraperitoneally with filter-sterilized 5 g/kg glucose in 0.9% NaCl following overnight fasting. A tail vein blood sample was taken before the injection and 30, 60, 90, and 120 min after the injection for determination of blood glucose. The insulin tolerance test (ITT) was performed by the UVA Animal Characterization Core following their approved protocols and as described by others (8). Briefly, the ITT was performed in random-fed mice starting at 2:00 P.M. The mice were injected intraperitoneally with insulin (0.75 U/kg) in 0.9% NaCl. A tail vein blood sample was taken immediately before and 15, 30, 45, and 60 min after the injection for determination of blood glucose levels. GTT and ITT were performed on different sets of mice to prevent unnecessary repetitive stress that could interfere with body weight gain. GTTs and ITTs were assessed by two-way ANOVA for time- and treatment-dependent differences between groups.

Serum lipid, cytokine, and adipokine profiles.

Blood samples from random-fed mice were obtained by cardiac puncture after killing the mice by cervical dislocation. The samples were centrifuged to isolate the blood serum, which was analyzed for a lipid profile in the UVA Clinical Pathology Core. Cytokine and adipokine levels were assessed using enzyme-linked immunosorbent assay immunoassay kits (ALPCO Diagnostics, Salem, NH). Intra-assay variation was <10%.

Islet treatment and insulin secretion.

Mouse pancreatic islets were isolated as described previously (69). After overnight incubation, islets were tested for insulin secretion as described previously (12, 68, 69). Briefly, islets were preincubated at 37° and 5% CO₂ for 1 h in a standard Krebs-Ringer-bicarbonate (KRB) solution, then washed and incubated in KRB supplemented with 3 mM glucose for 1 h followed by a 1-h treatment with KRB containing 28 mM glucose. The supernatant was collected after each treatment, and insulin concentration in the supernatant was measured by an enzyme immunoassay (EIA) method (ALPCO Diagnostic, Windham, NH) with a mouse insulin standard. Intra-assay variation was 6.6% and interassay variability <10%.

Immunohistochemistry.

After in vivo experiments were completed, mice were killed by cervical dislocation, and their pancreata and perigonadal visceral fat tissues were located and isolated. Tissue samples were fixed with 4% paraformaldehyde for 24–30 h at room temperature and embedded in paraffin (68). Five-micrometer-thick paraffin-embedded tissue sections were then deparaffinized and rehydrated in graduated alcohol in distilled water.

For visceral adipose tissue, antigens were retrieved using a high-temperature antigen-unmasking technique (Antigen Unmasking Solution; Vector Laboratories). The endogenous peroxidase was quenched using 0.5% H₂O₂ in methanol (Fisher Scientific) for 30 min at room temperature. The sections were then incubated for 30 min at room temperature with diluted normal blocking serum (Vector Laboratories) and stained at 4°C overnight with different primary antibodies, Mac2 (catalog no. AC18942; Accurate Chemical) and MCP-1 (catalog no. SC 1785; Santa Cruz Biotechnology), followed by a biotinylated secondary antibody (Vector Laboratories) for 60 min at room temperature and an avidin-biotin-peroxidase complex (Vector Laboratories) for 30 min at room temperature. Finally, the sections were developed with a diaminobenzidine substrate kit (Dako, Carpinteria, CA) and counterstained using hematoxylin. The expressions of MCP-1 and 12-LO were determined by microscopic observation of the diaminobenzidine reaction product on the analyzed sections. To assess adipose cell size in tissue sections, we measured cross-sectional cell areas from 40–50 cells for each mouse using Scion Image software. Data for adipose cell areas are presented in micrometers squared (μm²).

For islets, tissue sections of 5 μm thickness were treated with rabbit polyclonal insulin antibody (1:50) and goat polyclonal glucagon antibody (1:50) and stained using a secondary antibody linked to horseradish peroxidase as previously described (68). Islet size was quantified using ImageJ (NIH website: http://rsb.info.nih.gov/ij/download.html) to measure individual islet areas in representative images taken at ×400 magnification. To detect islet apoptosis, Cy2-conjugated goat anti-rabbit and Cy3-conjugated donkey anti-mouse antibodies (Jackson Immunoresearch Laboratories, West Grove, PA) were used as fluorescent markers for caspase-3 and insulin, respectively. Pancreatic tissue sections were submerged in 0.1% antigen unmasking solution (Vector Laboratories, Burlingame, CA) and microwaved for 20 min while replacing evaporated buffer with distilled water after 5, 9, 13, 17, and 19 min to prevent tissue dehydration. Samples were cooled at room temperature for 1 h, washed in PBS, and treated with 2.5% nonimmune serum for 30 min to reduce nonspecific staining. Tissue sections were incubated overnight at 4°C in rabbit polyclonal anti-active caspase-3 antiserum (1:100) or mouse monoclonal anti-insulin antiserum (1:100). Samples were washed in PBS and incubated for 1 h at room temperature in Cy2-conjugated goat anti-rabbit (1:50) secondary antibody to detect caspase-3 (in green) or Cy3-conjugated donkey anti-mouse (1:200) secondary antibody to detect insulin (in red). Negative controls were prepared by substituting either goat or rabbit serum for the primary antibodies. Slides were washed in PBS and examined for fluorescence using a Zeiss Axioplan microscope. The excitation wavelengths were selected using bandpass filters 450–515 nm and 546–590 nm for Cy2 and Cy3, respectively. Images were obtained with a SPOT CCD camera (Diagnostic Instruments, Sterling Heights, MI) and analyzed using SPOT software. Similarly sized islets were assessed for these studies, and caspase-3 staining was normalized to total nuclei/islet for a subset of islets from each treatment group. Total nuclei/islet did not differ among groups (P > 0.20).

12-LO activity measured by 12(S)-HETE production.

12(S)-HETE produced from islets was measured using an EIA (Assay Designs, Ann Arbor, MI) as previously described (13). Briefly, 12(S)-HETE was extracted from islets using 200 mM methanolic sodium hydroxide solution supplemented with 40 mM n-propyl gallate, acidified to pH 3–3.5, and purified by C-18 bond elution column (Varian Associates, Harbor City, CA). Purified samples were further subjected to HPLC purification (36, 51) and then to an EIA using a specific 12(S)-HETE antibody.

Isolation of stromal vascular cells and flow cytometry.

Samples of adipose tissue were digested with collagenase for isolation of stromal vascular cells (SVC). Briefly, adipose tissue was isolated and minced into fine pieces. Minced samples were placed in 5 ml HEPES-buffered DMEM containing 10 mg/ml fatty-free BSA and centrifuged at 500 g for 5 min to remove debris, erythrocytes, and other blood cells. Collagenase (Sigma-Aldrich) was added to 1 mg/ml before incubation at 37°C for 1 h on an orbital shaker. Once digestion was complete, the cell suspension was filtered sequentially through 100- and 40-μm filters and spun at 500 g for 5 min. The pellet containing SVC was resuspended in erythrocyte lysis buffer and incubated for 5 min. The erythrocyte-depleted SVC were centrifuged at 500 g for 5 min, and the pellet was resuspended in FACS buffer. The SVC were counted using a hemocytometer. The SVC were incubated in the dark with fluorophore-conjugated primary antibodies for 30 min at room temperature. Antibodies used in these studies included the following: CD-11b-FITC (Serotec, catalog no. MCA711F), CD45-PerCP (BD Pharmingen, catalog no. 557235), and F4/80-APC (Serotec, catalog no. MCA497APC). Cells were washed two times with 2 ml FACS buffer and analyzed on a FACSCalibur. Analysis was performed using CellQuest software. The macrophage content was defined by both F4/80 and CD11b positive in CD45 positive gate. The macrophages were determined per gram of starting adipose tissue used as starting material.

Statistics.

Statistical analysis was performed using Graph Pad Prism version 4.03 software. All data are presented as mean values ± SE, unless otherwise stated. One-way ANOVA followed by a Tukey posttest was used for comparing more than all groups unless otherwise stated. A P value <0.05 was used to indicate statistical significance.

RESULTS

Metabolic effects of western diet on B6 and 12-LOKO mice.

Mice from both the B6 and 12-LOKO groups were placed on either a chow or western diet and monitored weekly for body weight and blood glucose under nonfasting conditions. Multiple trials were performed under the same experimental conditions, producing similar findings in each case (see Table 1 for three combined trials of 8 wk on diet). Blood glucose measurements under fed conditions remained <250 mg/dl for all trials in all treatment groups, indicating no incidence of diabetes. Although there was a trend toward increased fed blood glucose levels for both strains on the high-fat diet (see Table 1), this effect was not statistically significant (P > 0.05). In contrast, fasting blood glucose readings at 8 wk on the diet showed significant increases in western-fed B6, but not western-fed 12-LOKO mice, compared with chow-fed controls, consistent with developing insulin resistance. Fasting insulin levels were not available from the mice in these studies; however, we obtained fasting serum insulin from mice in parallel studies under identical conditions following 12 wk on diet (Table 1). Chow-fed mice from both B6 and 12-LOKO mice showed similar fasting insulin levels. Western-fed B6 mice showed a substantial increase in fasting insulin compared with the chow-fed groups, whereas western-fed 12-LOKO mice did not.

Table 1.

Blood glucose, body weight, and serum lipid profile

	B6 Chow	B6 West	12-LOKO Chow	12-LOKO West
Body weight at start, g	23.7±0.8	22.4±0.7	23.7±0.9	24.9±0.8
Body weight at 8 wk, g	28.6±0.8	34.2±0.8*	26.3±0.7	32.1±0.6*
Fed blood glucose at 8 wk, mg/dl	128±7	151±8	130±7	155±9
Fasting blood glucose at 8 wk, mg/dl	106±7	163±10*	113±10	130±10
Fasting insulin at 12 wk, pg/ml	234±40	518±74*	266±33	359±73
Total cholesterol	60±2.3	107±7.9*	57±3.1	113±8.1*
Triglycerides	46±1.4	129±6.9*	42±2.3	125±7.6*
LDL-C	2±0.21	43±1.7*	3±0.78	45±1.4*
HDL-C	50±1.9	46±2.1	49±2.3	52±1.7

Data represent averages of 3 separate trials with 4–6 mice/trial. Western diet increased total cholesterol (TC), triglycerides (TG), and low-density lipids (LDL-C) for both C57BL/6 (B6) and 12-lipoxygenase-null (12-LOKO) male compared with chow-fed controls. No change in high-density lipids (HDL-C) was observed among the groups. Lipid profiles for the western-fed B6 and western-fed 12-LOKO groups did not differ significantly from one another.

^*(P < 0.05). Fasting insulin measurements were made from different mice in parallel studies (n = 15–21 mice/group).

Mean body weights at the end of the 8-wk study for western-fed B6 and western-fed 12-LOKO mice were significantly greater than for either chow-fed group (P < 0.01), as expected. Western-fed groups did not differ from one another in body weight, nor did chow-fed groups differ (P > 0.05, n > 15 mice/group). Similar findings were observed at 24 wk (B6 chow, 35.6 ± 4.7; B6 western, 47.4 ± 3.9; 12-LOKO chow, 34.5 ± 4.0; 12-LOKO western, 43.4 ± 4.7 g, n = 7–9 mice/treatment group). When comparing changes in body weight directly, however, western-fed B6 mice gained 3.5 g more than western-fed 12-LOKO mice (P < 0.05) by the end of the 8-wk diet, suggesting that 12-LO deficiency slightly reduces the rate of weight gain. Note, however, adipocyte cell size did not differ between 12-LOKO and B6 mice on the western diet (see Fig. 6C).

We also compared lipid profiles at 8 wk to determine if the western diet had effectively altered blood serum cholesterol and triglyceride levels. As shown in Table 1, low-density lipoproteins, triglycerides, and total cholesterol were substantially elevated in western-fed mice in both the B6 and 12-LOKO groups. Western diet had no effect on high-density lipoproteins in either group. Overall, western diet effects were thus similar in both B6 and 12-LOKO groups for these measures.

Glucose and insulin tolerance.

Mice were given a GTT and ITT in week 7, a week before the study termination. As shown in Fig. 1A, B6 chow, 12-LOKO chow, and 12-LOKO western-fed mice groups displayed very similar responses to the glucose challenge: at 30 min blood glucose peaked in response to the glucose bolus and decreased steadily to ∼200 mg/dl. In contrast, the B6 western-fed mice began with higher fasting blood glucose before the glucose bolus (P < 0.01; see also Table 1), and glucose levels remained elevated following the glucose injection compared with the other three groups (P < 0.01), suggesting that these mice have impaired glucose tolerance. Similarly, the ITT results indicate that western-fed B6 mice have reduced insulin sensitivity compared with the other treatment groups (Fig. 1B). These findings thus suggest that 12-LOKO mice maintain normal glucose tolerance and insulin sensitivity under the western diet.

Fig. 1.

Western-fed C57BL/6 (B6) mice have reduced glucose and insulin tolerance. A: glucose tolerance test (GTT). Blood glucose levels were measured at 30-min intervals following an ip glucose bolus. Western-fed B6 mice showed a higher fasting blood glucose level before glucose stimulation and also a slower rate of glucose clearance compared with other treatment groups. B: insulin tolerance test (ITT). Blood glucose levels were measured at 30-min intervals following an insulin injection. Western-fed B6 mice showed a reduced and slower response to insulin compared with other treatment groups. The western-fed B6 mice differed from the other groups in GTT and ITT as determined by two-way ANOVA for time and treatment dependence (P < 0.05); n = 4–6 mice/treatment group.

Western diet alters serum cytokines and adipokines in B6 mice, but not 12-LOKO mice.

We also obtained blood serum measurements of key circulating cytokines from the 8-wk-diet trials. We chose to specifically examine adiponectin, tumor necrosis factor-α (TNF-α), and interleukin-6 (IL-6), because these adipocyte-derived factors are associated with western-diet-induced inflammation and metabolic changes. As shown in Fig. 2, adiponectin, which is considered to protect against inflammatory damage, was reduced in western-fed B6 mice compared with chow-fed mice. In contrast, the proinflammatory cytokines TNF-α and IL-6 were increased in western-fed B6 mice. Levels of these cytokines in western-fed 12-LOKO mice, however, were similar to chow-fed mice, suggesting that western-fed 12-LOKO mice are protected against increases in proinflammatory cytokines and reductions in adiponectin.

Fig. 2.

Western diet alters serum adipokine levels in B6 mice, but not in 12-lipoxygenase-null (12-LOKO) mice. Western-fed B6 mice showed reduced adiponectin, increased tumor necrosis factor (TNF)-α, and increased interleukin (IL)-6 in blood serum compared with other treatment groups. *P < 0.05; n = 4–6 mice/treatment group.

Effects of western diet on pancreatic islets.

After completing in vivo measurements, mice were killed, and pancreatic islets were isolated to investigate islet function and histology. We first examined 12-LO activity in pancreatic islets. Using a sensitive HPLC system to measure the 12-LO product 12-HETE labeled with carbon-14, we investigated whether increased 12-LO activity in islets is associated with the effects of western diet on B6 mice. As shown in Fig. 3A, levels of the 12-LO product 12-HETE were significantly increased by western diet in B6 mice but were barely detectable in the 12-LOKO mice on either diet. Therefore, western diet increases 12-LO activity in islets from B6 mice. Consistent with the absence of 12-LO and other enzyme activities compensating for the lack thereof, 12-HETE was barely detectable in islets of 12-LOKO mice.

Fig. 3.

Islet 12-lipoxygenase (12-LO) activity and glucose-stimulated insulin secretion. A: 12-LO activity is enhanced by western diet in B6 mice (*P < 0.05) and greatly reduced in 12-LOKO mice (#P < 0.05), as measured by HPLC of the 12-LO product 12-HETE labeled with carbon-14. B: islets from western-fed B6 mice showed a reduced glucose stimulation index compared with other groups; n = 8–10 mice/treatment group using 50 islets/mouse to measure insulin.

Islets isolated in two trials were assessed for physiological function by measuring basal insulin secretion (3 mM glucose) followed by glucose-stimulated insulin secretion (28 mM glucose). Result for insulin secretion in 3 mM (basal) glucose are as follows: B6 chow, 5.7 ± 1.3; B6 western, 19.9 ± 5.2 (P < 0.05); 12-LOKO chow, 7.4 ± 2.3; 12-LOKO western, 8.5 ± 2.9 pg/ml. Thus western diet led to a significant increase in basal insulin secretion in B6 mice but not 12-LOKO mice. Insulin secretion in 28 mM glucose conditions was generally higher in islets from western-fed mice although the differences were not significant (B6 chow, 11.3 ± 4.1; B6 western, 22.9 ± 7.6; 12-LOKO chow, 11.6 ± 2.5; 12-LOKO western, 22.7 ± 6.6 pg/ml). As shown in Fig. 3B, the ratio of stimulated to basal glucose (glucose stimulation index) was thus reduced for the B6 western-fed mice compared with the other treatment groups. These results suggest a loss of glucose responsiveness in islets from B6 western-fed mice, but not in 12-LOKO western-fed mice.

Pancreata were also harvested and sectioned to examine islet histology. As shown in Fig. 4, islets from B6 mice on a western diet were significantly larger than islets from other groups. Examples of typical islets from each treatment group stained for insulin with horseradish peroxidase are shown in Fig. 4A. Mean islet area shown in Fig. 4B was calculated using pancreatic sections from four or more mice for each treatment group. The results demonstrate increased islet area in western-fed B6 mice compared with the chow-fed B6 mice. The 12-LOKO mice on a western diet showed no increase in islet size, suggesting that, despite similar weight and lipid profile, this compensatory mechanism was not triggered in 12-LOKO mice.

Fig. 4.

Western-fed B6 males typically had larger pancreatic islets. A: representative examples of islets stained with horseradish peroxidase (HRP) for insulin were taken from pancreatic tissue. B: mean islet areas estimated by pixel count and converted to μm for all treatment groups. The areas of at least 60 islets were assessed for each treatment group from at least 3 mice/treatment group. *P < 0.05, differences among groups.

Because inflammatory damage can result in β-cell death, we examined a possible role of 12-LO in β-cell apoptosis. We hypothesized that western-diet-induced β-cell stress would be reduced in 12-LO-deficient mice, whereas islets from B6 mice on the western diet would show an increase in apoptotic nuclei measured by caspase-3 staining. As shown in the representative examples in Fig. 5A, adjacent pancreatic sections stained for insulin (red) or caspase-3 (green) show islets from 12-LOKO mice were protected from western-diet-induced increases in cell death that occurred in islets from B6 mice on western diet. As summarized in Fig. 5B, the number of caspase-3-positive nuclei observed per islet in western-fed B6 mice was also substantially greater than for any other treatment group, which correlated with the increased 12-LO activity also observed in western-fed B6 mice. Furthermore, the mean number of caspase-3-positive cells in islets from 12-LOKO mice on either diet was significantly lower than in islets of chow-fed B6 mice (P < 0.02), which is consistent with the hypothesis that reducing 12-LO prevents damage that in turn causes apoptosis.

Fig. 5.

Apoptosis is increased among western-fed B6 mice but not 12-LOKO mice. A: representative examples of islets stained for the apoptotic marker caspase-3 (green) and for insulin (red) in adjacent 5-μm pancreatic sections. B: mean detection of apoptosis as determined by caspase-3-positive nuclei/islet, assessing 12–18 islets/treatment group using pancreatic sections from 3–4 mice/treatment group. *P < 0.05 vs. B6 chow group.

Inflammatory effects of western diet in adipose tissue.

We also looked at inflammatory effects of western diet in adipose tissue. In studies of 12 and 24 wk of duration on diet, visceral adipose tissue was isolated, and fluorescence activated cell sorting was used to count macrophages in the adipose tissue. At both 12 (Fig. 6A) and 24 (Fig. 6B) wk, B6 and 12-LOKO mice on western diet showed progressively increased numbers of macrophages per gram of adipose tissue compared with chow-fed controls. However, macrophage content was significantly less in western-fed 12-LOKO mice compared with western-fed B6 mice, supporting the hypothesis that 12-LO promotes inflammation in visceral fat.

Fig. 6.

Macrophage counts in adipose tissue. A and B: macrophages identified by cell sorting and counted per gram of adipose tissue for B6 and 12-LOKO mice on either chow or western diet for 12 wk (A) and 24 wk (B). C: mean adipose cell area in pixels for each treatment group (n = 4 mice/group, n = 40–50 cells/mouse). P < 0.05 vs. B6 chow group (*) and vs. B6 western-fed group (#).

Western diet leads to enlargement of adipocytes, which is associated with impaired adipocyte function and increased apoptosis. Because this, in turn, triggers an inflammatory response causing increased macrophage infiltration (25), we investigated adipocytes for differences in size. As shown in Fig. 6C, western diet significantly increased adipose cell size in both B6 and 12-LOKO mice. Mean adipose cell areas from 12-LOKO and B6 mice did not differ in size. We also examined the frequency distribution of adipocyte sizes among western-fed mice by distributing adipocytes in intervals of 2,000 μm². A trend toward more adipocytes of size <2,000 μm² was observed among western-fed 12-LOKO mice by Chi Square test (P = 0.07). The larger number of very small cells may help explain the larger error bar for the 12-LO KO mice in Fig. 6C. No differences were observed in other adipocyte size intervals between western-diet-fed strains (P > 0.19). Together with the data demonstrating decreased macrophage infiltration (Fig. 6, A and B) these findings suggest that, in response to the western diet, 12-LO-deficient adipocytes, despite increased fat storage, maintain an environment that is not proinflammatory and proapoptotic and consequently does not attract macrophages.

Differences in inflammation were further evaluated by immunohistochemical staining of adipose tissue for inflammatory markers after 24 wk on diet. Mac-2 was used as a marker for leukocytes. As shown in Fig. 7B, representative of observations from 7–8 mice/group, mac-2 staining was observed in pockets throughout the adipose tissue of western-fed B6 mice. In contrast, very little mac-2 staining was evident in the chow-fed B6 mice (Fig. 7A) or western-fed 12-LOKO mice (Fig. 7C). The staining was quantified in Fig. 7D, confirming a greater presence of mac-2 staining in western-fed B6 mice compared with western-fed 12-LOKO mice.

Fig. 7.

Macrophage staining in adipose tissue. Staining for leukocytes (mac-2) in adipose tissue from B6 mice on a chow diet (A), B6 mice on western diet (B), and 12-LOKO mice on western diet (C) for 24 wk. D: %cells showing mac-2-positive staining in adipose tissue. Data are representative of 7–8 mice/treatment group. *P < 0.05, difference with B6 chow group.

Because we hypothesized that the absence of 12-LO may prevent the attraction of macrophages to adipose tissue resulting in reduced inflammation, we evaluated the content of MCP-1, a known signaling molecule to attract monocytes to sites of inflammation. As shown in Fig. 8, A and B, visceral adipose tissue from B6 mice on western diet showed clear MCP-1 staining (brown) around many more adipocytes (Fig. 8A) than adipose tissue from 12-LOKO mice (Fig. 8B). The boxed portions of Fig. 8, A and B, are magnified in Fig. 8, C and D, respectively, to show greater resolution. These findings suggest that reduced expression of MCP-1 in 12-LOKO mice could result in reduced macrophage infiltration in visceral fat.

Fig. 8.

Monocyte chemoattractant protein (MCP)-1 staining in adipose tissue. A and B: staining for MCP-1 in adipose tissue from B6 (A) and 12-LOKO (B) mice on western diet for 24 wk. Boxed areas in A and B are shown at higher magnification in C and D, respectively. Data are representative of staining from 7–8 mice/treatment group.

DISCUSSION

This study shows for the first time that the 12-LO-deficient mouse is protected against western diet-induced metabolic and inflammatory changes. We replicated many adverse effects of the western diet previously observed in rodents, including increased body weight, increased proinflammatory cytokines and reduced adiponectin levels (3, 26, 29), altered serum lipid profile (9), impaired glucose and insulin tolerance (53, 57), increased macrophage infiltration in visceral fat concomitant with increased adipocyte size (62, 66, 67), reduced glucose-stimulated insulin secretion in islets (1, 34), and enlarged islet size (47, 60, 61). Although western-fed 12-LOKO mice showed a similar marked change in several of these endpoints, glucose and insulin tolerance remained normal, isolated islets functioned normally, and macrophage infiltration in visceral fat was much lower than in western-fed B6 mice.

Inflammation of adipose tissue in obesity mediates the development and exacerbation of insulin resistance and metabolic defects (17, 22, 67). Normal levels of proinflammatory cytokines (TNF-α and IL-6) and anti-inflammatory adiponectin in the western-fed 12-LOKO mice provide evidence for a role of 12-LO in the inflammatory response to a western diet. These adipose-tissue-derived factors are increased (TNF-α and IL-6) and decreased (adiponectin) in rodents fed high-fat diets (3, 26, 29). We have previously reported that 12-LO products can directly induce the expression and release of proinflammatory cytokines from macrophages (60) and adipocytes (38). Because adiponectin has been shown to decrease insulin resistance, whereas TNF-α and IL-6 increase insulin resistance (3, 29), reducing 12-LO may prevent the western diet-induced insulin resistance by decreasing the inflammatory response in visceral adipose tissue.

The inflammation of adipose tissue in obesity is also marked by macrophage infiltration (17, 22, 67). Western-diet-fed 12-LOKO mice showed significantly reduced macrophage infiltration and MCP-1 expression in visceral fat, suggesting that 12-LO function contributes to macrophage recruitment. An enlargement of adipocyte size is considered to be the most important factor in creating a proinflammatory environment in adipose tissue (17, 22, 67). Adipocyte size did not differ between western-diet-fed B6 and 12-LOKO mice, providing further support for the hypothesis that 12-LO plays a primary role in promoting inflammation in adipose tissue. However, we observed a difference in body weight gain between western-fed 12-LOKO and B6 mice (3.5 g greater in B6 mice). This may reflect a difference in fat mass, which was not determined in this study. Thus we cannot exclude the possibility that a difference in visceral fat mass may have contributed to the reduced inflammation and improved metabolic phenotype observed in the 12-LOKO mice. Nevertheless, it is unlikely that this alone could explain all of the protective effects observed in the western-fed 12-LOKO mice, particularly at the level of the adipocyte.

Reduced inflammation and consequently improved insulin sensitivity in 12-LOKO mice may at least partially explain the observed effects on size and insulin secretion in islets. Obesity, hyperglycemia, and insulin resistance are known to trigger islet mass expansion to promote increased insulin production (19, 57, 58, 60, 61). As insulin resistance increases, islets also compensate by shifting glucose sensitivity to secrete more insulin at lower glucose concentrations. Glucose-stimulated insulin secretion among islets isolated from western-fed 12-LOKO mice was similar to chow-fed controls in this study, whereas basal insulin secretion was increased and the overall glucose stimulation index reduced among islets from western-fed B6 mice. These findings are consistent with known effects of western diet, both in terms of decreased glucose and insulin tolerance in vivo (53, 57) and in the islet in vitro (1, 15, 34, 60). Because 12-LOKO mice did not show evidence of insulin resistance, islets from western-fed 12-LOKO mice thus also did not need to adjust to increasing workload and did not show significant changes in glucose stimulation or islet size.

12-LO deficiency may also protect islets from damage in a more direct manner. Previous studies indicate that 12-LO products can induce islet dysfunction and death in vitro (13, 48). Also, increased rates of cell death in β-cell lines overexpressing 12-LO (48) and in isolated islets from mice that overexpress 12-LO (44) have been observed. The results in the current study support a role of 12-LO in increasing cell death in islets under stressful conditions. The caspase-3 staining in pancreatic tissue demonstrated a reduction in islet cell apoptosis in 12-LOKO mice, and this effect was observed in both chow and western-fed 12-LOKO mice. For the western-fed mice, this difference may be explained by the difference in insulin sensitivity and inflammation in vivo. However, for the chow-fed mice, the differences in cell death may be explained by protection of 12-LOKO islets from the stressful effects during islet isolations.

As discussed above, 12-LO deficiency in 12-LOKO mice may provide protection from the western diet-induced changes by actions in several different tissues. 12-LO is normally expressed at low levels in many tissues, including macrophages, adipocytes, and pancreatic β-cells (4, 16, 42). The inflammatory effects of 12-LO may occur with higher levels of expression or activity, as induced by western diet. Western diet increases 12-LO expression in adipocytes in B6 mice as we have recently reported (38) and in islets as we demonstrate in the present study. Our study with the global 12-LOKO mice thus provides evidence for a possible role of 12-LO in western diet-induced changes in adipose tissue and in islets. Because of the global nature of the 12-LO deletion, reduction of 12-LO in islets, muscle, fat, macrophages, and other tissues may all contribute to the phenotype reported in this study (23, 28, 41, 45, 63). The complex interactions between and among these tissues in vivo preclude definitive answers regarding the site(s) of 12-LO action with the present model even when studying these tissues in vitro. To determine the precise role of 12-LO in islets, adipocytes, and other cells and tissues, tissue-specific and inducible knockout models will need to be generated and characterized.

In summary, our findings suggest a key role of 12-LO in the development of inflammation and metabolic derangements seen with western diet feeding. We propose that molecular or pharmacological reduction of 12-LO expression or activity could be a new therapeutic target to improve the metabolic state and inflammatory damage associated with high-fat intake and visceral adiposity.

GRANTS

This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases RO1 Award DK-55240 to J. L. Nadler and Grants DK-063609 and DK-063609, and HL-P01-55798 (J. L. Nadler).