Obesity-associated improvements in metabolic profile through expansion of adipose tissue

Abstract

Excess caloric intake can lead to insulin resistance. The underlying reasons are complex but likely related to ectopic lipid deposition in nonadipose tissue. We hypothesized that the inability to appropriately expand subcutaneous adipose tissue may be an underlying reason for insulin resistance and beta cell failure. Mice lacking leptin while overexpressing adiponectin showed normalized glucose and insulin levels and dramatically improved glucose as well as positively affected serum triglyceride levels. Therefore, modestly increasing the levels of circulating full-length adiponectin completely rescued the diabetic phenotype in ob/ob mice. They displayed increased expression of PPARgamma target genes and a reduction in macrophage infiltration in adipose tissue and systemic inflammation. As a result, the transgenic mice were morbidly obese, with significantly higher levels of adipose tissue than their ob/ob littermates, leading to an interesting dichotomy of increased fat mass associated with improvement in insulin sensitivity. Based on these data, we propose that adiponectin acts as a peripheral "starvation" signal promoting the storage of triglycerides preferentially in adipose tissue. As a consequence, reduced triglyceride levels in the liver and muscle convey improved systemic insulin sensitivity. These mice therefore represent what we believe is a novel model of morbid obesity associated with an improved metabolic profile.

Figure 1. ob/ob mice overexpressing adiponectin display physiological overexpression of adiponectin in circulation with normal adiponectin complex distribution.

(A) Circulating adiponectin levels were measured by RIA in 12-week-old adiponectin transgenic ob/ob mice (Ad Tg ob/ob), their ob/ob littermates, and ob/ob mice with PPARγ agonist treatment (n = 5 mice/group). (B) Tissue levels of adiponectin were measured in various fat pads, determined by Western blot analysis, and normalized with GDI in 12-week-old male adiponectin transgenic ob/ob mice, their ob/ob littermates, and ob/ob mice with PPARγ agonist treatment (n = 5 mice/group). (C) The complex distribution of adiponectin of female mice was determined using gel filtration chromatography of serum samples followed by Western blot analysis of adiponectin in different fractions. A representative sample is shown. HMW, high molecular weight form; LMW, low molecular weight form (hexamer). *P < 0.05; **P < 0.01 by Student’s t test.

Figure 2. Adiponectin improves the metabolic profile of ob/ob mice.

(A) Circulating glucose and insulin levels measured during OGTT. (B) Plasma triglyceride (TG) levels measured during lipid challenge in adiponectin transgenic ob/ob mice and their ob/ob littermates. (C) Plasma triglyceride levels were measured during a lipid challenge in adiponectin transgenic ob/ob mice after an injection of 2 monoclonal anti-adiponectin antibodies (Mono-Ab) or control mouse IgGs (CTRL-IgG). (D) Plasma insulin (right) and FFA levels (left) monitored after i.p. CL-316,243 injection in adiponectin transgenic ob/ob mice and their ob/ob littermates. (E) Cholesterol and triglyceride levels were measured in plasma samples subfractionated by gel filtration chromatography. IDL, intermediate density lipoprotein. (F) apoE levels were measured in the same fractions by Western blot analysis (top) and quantitated (bottom). (G) Adiponectin transgenic ob/ob mice and their ob/ob littermates were fed normal chow (N chow) or high-fat diet (HF) for 6 weeks. Fasting glucose levels (top left), glucose levels during an OGTT (top right), insulin levels (bottom left), and FFA levels (bottom right) were determined in adiponectin transgenic ob/ob mice and their ob/ob littermates (A–G, n = 5 mice/group). (H) 2DG was centrally delivered to adiponectin transgenic ob/ob mice, ob/ob littermates, and WT mice, and blood glucose levels were monitored during indicated times (n = 3 mice/group). Experiments were performed with 10- to 12-week-old adiponectin transgenic ob/ob mice, ob/ob littermates, and WT male mice. *P < 0.05; **P < 0.01. Panels A and B were analyzed by ANOVA; panels D and G were analyzed by Student’s t test.

Figure 3. ob/ob mice expressing adiponectin are heavier but healthier than ob/ob mice, displaying increased fat mass and decreased liver triglycerides.

(A) Changes in body weight were monitored (n = 15 mice/group). Representative animals are shown: an ob/ob mouse and an adiponectin transgenic ob/ob littermate at 12 weeks. (B) Body composition of 5-week-old mice was measured by MRI (n = 3 mice/group). (C) Body weight gain was tracked on normal chow or on a HF diet in 10-week-old adiponectin transgenic ob/ob mice and ob/ob littermates. (D) Sections of gonadal WAT and (E) pancreatic islets from adiponectin transgenic ob/ob mice (top) and ob/ob littermates (bottom) were H&E stained. The average adiponectin size calculated (D, right panel) and surface area of >100 islets stained for insulin was measured and average islet area is indicated (E, right panel). (F) Frozen sections of liver from adiponectin transgenic ob/ob mice and ob/ob littermates were stained with oil red O (left panel). Liver triglyceride content (middle panel) and DAG levels (right panel) were analyzed. (G) Intraabdominal visceral fat pads from WT mice, adiponectin transgenic ob/ob mice, and ob/ob littermates were weighed and presented as percentage of total body weight (F and G, n = 4–5 mice/group). (H) Pericardial fat pads from 10-week-old female mice were imaged by MRI and reconstituted into a 3D representation. A representative example is shown. (I) Mice were fasted for 6 hours and injected with 1 mU of insulin/g body weight. At indicated time points, 1 mouse was sacrificed and liver protein was extracted and analyzed for the total and phosphorylated forms of the indicated targets using 3 independent measurements for each time point. *P < 0.05; **P < 0.01. Panels A and C were analyzed by ANOVA; panels D, F, G, and I were analyzed by Student’s t test.

Figure 4. Excess adiponectin normalizes pancreatic islet histology and improves both local and systemic inflammation.

(A) Immunofluorescence staining for insulin (green) and glucagon (red) was performed in pancreatic sections from 10-week-old male WT mice (top), ob/ob littermates (middle), and adiponectin transgenic ob/ob mice. (B) Immunohistochemical staining of gonadal adipose tissue with anti-F4/80 antibodies suggests reduced infiltration of macrophages. Note that the difference in average adipocyte size between the 2 genotypes is apparent but less pronounced at this younger age. (C) FACS analysis of the stromal vascular fraction using F4/80 antibodies (left) and qRT-PCR for F4/80 (right) (n = 10 mice/group) were performed in gonadal WAT of adiponectin transgenic ob/ob mice and ob/ob littermates. Black curve shows control IgG staining; red curve shows anti-F4/80 staining. (D) mRNA levels of vWF and VCAM-1 from WAT were measured by qRT-PCR. A ratio of VCAM-1/vWF is provided as an indicator of local inflammation in the vascular endothelium. (E) Circulating IL-6 was measured in adiponectin transgenic ob/ob mice and ob/ob littermates (top); adipose tissue TNF-α message levels (bottom). (C–E) For these studies, 14-week-old male mice were used (n = 10 mice/group). *P < 0.05. Panels C–E were analyzed by Student’s t test.

Figure 5. Increased adiponectin in ob/ob mice triggers improvements in lipid metabolism.

(A) LPL activity of 12-week-old mice was measured in a number of different fat pads as well as in plasma. (n = 5 mice/group). (B) mRNA levels of PEPCK, DGAT-1, PPARγ2, and aP2 were measured by qRT-PCR (n = 5 mice/group) in WAT from 12-week-old adiponectin transgenic ob/ob mice and ob/ob littermates. *P < 0.05. Panels A and B were analyzed by Student’s t test.

Figure 6. Metabolic cage studies on adiponectin-expressing ob/ob mice and ob/ob littermates under normal and fasting conditions.

(A) Daily food uptake was monitored at 6 weeks of age (left) and 20 weeks of age (middle and right) and represented in absolute amounts (left and middle) or expressed as food intake per total body weight (BW) (middle and right) in adiponectin transgenic ob/ob mice and ob/ob littermates (n = 3–5 mice/group). (B) Core body temperature (temp.), (C) activity (ambulatory on the left, rearing on the right), (D) VO₂, and (E) RER were measured in adiponectin transgenic ob/ob mice and ob/ob littermates. (n = 3 mice/group). Male adiponectin transgenic ob/ob mice and ob/ob littermates were denied access to food for 36 hours. (F) RER and (G) activity (ambulatory on the left, rearing on the right) were analyzed during the course of study. (H) Glucose (top) and FFAs (bottom). Blood was collected every 4 hours over a period of 36 hours to measure plasma glucose and FFAs. *P < 0.05; **P < 0.01. For B–H, 12-week-old male mice were used. All data were analyzed by ANOVA. BF, before food removal.

Figure 7. Pronounced hypoglycemia in female adiponectin transgenic ob/ob mice during fasting.

In a separate experiment, female and male adiponectin transgenic ob/ob mice and ob/ob littermates were examined during a fast. (A) VO₂ (left) and core body temperature (right) were monitored in 12-week-old male mice. (B) Change of body weight over the fasting period in 12-week-old female mice is shown. (C) Glucose levels of 12-week-old females during a 36-hour fast is shown. (D) Mice were sacrificed after 36 hours of fasting, and liver tissue was collected. qRT-PCR for PEPCK and G-6-P was performed from RNA of liver (A–D; n = 3 mice/group). *P < 0.05. All data were analyzed by ANOVA except panels A and D, which were analyzed by Student’s t test.