Vaccination against GIP for the treatment of obesity

Abstract

Background: According to the WHO, more than 1 billion people worldwide are overweight and at risk of developing chronic illnesses, including cardiovascular disease, type 2 diabetes, hypertension and stroke. Current therapies show limited efficacy and are often associated with unpleasant side-effect profiles, hence there is a medical need for new therapeutic interventions in the field of obesity. Gastric inhibitory peptide (GIP, also known as glucose-dependent insulinotropic polypeptide) has recently been postulated to link over-nutrition with obesity. In fact GIP receptor-deficient mice (GIPR(-/-)) were shown to be completely protected from diet-induced obesity. Thus, disrupting GIP signaling represents a promising novel therapeutic strategy for the treatment of obesity.

Methodology/principal findings: In order to block GIP signaling we chose an active vaccination approach using GIP peptides covalently attached to virus-like particles (VLP-GIP). Vaccination of mice with VLP-GIP induced high titers of specific antibodies and efficiently reduced body weight gain in animals fed a high fat diet. The reduction in body weight gain could be attributed to reduced accumulation of fat. Moreover, increased weight loss was observed in obese mice vaccinated with VLP-GIP. Importantly, despite the incretin action of GIP, VLP-GIP-treated mice did not show signs of glucose intolerance.

Conclusions/significance: This study shows that vaccination against GIP was safe and effective. Thus active vaccination may represent a novel, long-lasting treatment for obesity. However further preclinical safety/toxicology studies will be required before the therapeutic concept can be addressed in humans.

Figure 1. Vaccination against GIP results in high, specific antibody titers.

(A) Schematic diagram of GIP peptide coupling to Qβ VLPs. GIP peptides (aa 1–15), corresponding to the N-terminus of the active peptide, were covalently linked to Qβ VLPs via an SMPH linker. (B) GIP-specific antibody titers in vaccinated mice. Female mice were immunized s.c. with 100 µg of Qβ-GIP (days 0, 14, 28 and 42). Average GIP-specific antibody titers±SEM (n = 6) at the indicated time points are shown. (C) Cross-reactivity of GIP-specific antibodies. A serum pool from Qβ-GIP-immunized mice was incubated with increasing concentrations of GIP, GLP-1 or oxyntomodulin (OXM). The amount of free antibody was quantified by ELISA. (D) Recognition of the N-terminus of full length GIP. A serum pool from Qβ-GIP-immunized mice was preincubated either with GIP1–15 or GIP3–15 and free antibodies quantified by ELISA. (E) Sequestration of GIP in vivo. Qβ-GIP immunized mice or naïve mice were challenged i.v. with 1 ng of I¹²⁵-GIP. 30 minutes later the amount of antibody-bound GIP was determined. The percentage of antibody bound GIP±SEM (n = 4) is shown. (F) Antibody mediated blocking of GIP binding to its receptor. I¹²⁵-GIP was incubated with purified total IgG from Qβ-GIP or Qβ immunized mice and added to CHOK1-GIPR cells and bound GIP determined after an overnight incubation at 4°C. The final concentration of GIP was 20 ng/ml and the concentration of total IgG is shown on the x-axis. Error bars represent standard deviations from triplicates.

Figure 2. Vaccination against GIP protects against diet-induced obesity.

(A) Body weight gain in immunized mice. Female mice were immunized (days 0, 14, 28, 42 and 133) with 100 µg of Qβ-GIP or Qβ VLPs and placed on a high fat diet (35% fat w/v). The average body weight+/−SEM (n = 6) is shown. Body weight gain was was significanty reduced in Qβ-GIP- compared to Qβ VLP-immunized animals from day 70 onwards (two way ANOVA F(1,80) = 18.55, p<0.0001). (B) Body composition of mice shown in (A) was measured by DEXAscan on day 142. Average total body mass, lean and fat tissue mass+/−SEM (n = 6) are shown. A significant reduction in fat content (p = 0.01) was observed between the Qβ-GIP- and Qβ-vaccinated group as determined by t-test. DEXAscan images of one representative animal per group are shown. (C) Macroscopic analysis of mice after 142 days of treatment with Qβ-GIP or Qβ VLPs. A representative mouse from each group from the experiment described in Figure 3A–B is shown. (D) Body weight gain in immunized mice on a standard rodent diet. Female mice were immunized (days 0, 14, 28, 42 and 112) with Qβ-GIP or Qβ VLPs and fed a standard diet (4% fat w/v) during the whole experiment. Average body weights+/−SEM (n = 5) are shown. No significant difference between the two experimental groups was observed as determined (two way ANOVA F(1/88) = 0.81, p = 0.6751).

Figure 3. Vaccination against GIP increases energy expenditure and metabolic rate.

Indirect calorimetry in immunized mice. Female mice were immunized (days 0, 14, 28, 42 and 125) with Qβ-GIP (n = 8) or Qβ VLPs (n = 10) and placed on a high fat diet. Indirect calorimetry was performed on half of the group on day 128 and on the other half on day 139. Combined data from these measurements are shown. (A) Oxygen consumption (VO2). The left panel shows average oxygen consumption+/−SEM. Qβ-GIP-vaccinated animals display statistically, significantly increased VO2 (p<0.0001) over the 24 h period. Average oxygen consumption+/−SEM during the dark and light phase is shown on the right. VO2 was significantly increased in Qβ-GIP-vaccinated animals compared to Qβ controls in both the dark (p = 0.02) and light phase (p = 0.02). (B) Resting metabolic rate (RMR). RMR was increased in Qβ-GIP-vaccinated animals compared to Qβ VLP controls (p = 0.05). (C) Physical activity was determined by measuring beam brakes over a 24 h period. No significant differences were observed between the two experimental groups. (D) Respiratory quotient (RQ) was measured for 24 hours during the dark and light phase. Average RQ±SEM is shown. RQ is defined as VCO2 (L)/VO2 (L). The difference observed between the two experimental groups did not reach statistical significance. (E) Food intake. Food intake was monitored over three consecutive day after the energy expenditure experiment. Average daily food intake in mg/g body weight+/−SEM (n = 5 are shown). No statistically significant difference was observed between the experimental groups (p>0.05). All statistical analyses were performed by two-sided t-tests.

Figure 4. Vaccination against GIP does not alter glucose homeostasis.

(A) Non-fasted blood glucose levels. Female mice were immunized (days 0, 14, 28, 42 and 133) with Qβ-GIP or Qβ VLPs and placed on a high fat diet and non-fasted glucose levels were measured at the indicated time points. Average blood glucose level per group±SEM (n = 10) are shown. No significant difference between vaccinated and control group was observed (two way ANOVA F(1, 72) = 0.06, p = 0.8094) (B) Fasted blood glucose levels. Female mice were immunized (days 0, 14, 28, 42) with Qβ-GIP or Qβ VLPs and placed on a high fat diet. After a 16 h fast, blood glucose levels were measured in the afternoon. Average blood glucose level per group±SEM (n = 5) are shown. No significant difference between vaccinated and control group was (two way ANOVA F(1, 24) = 0.21, p = 0.6553) (C) Average fructosamine concentrations±SEM (n = 6) from mice in Figure 2A are shown. No significant difference between the two experimental groups was observed (two way ANOVA F(1,70) = 1.49, p = 0.25) (D) Average HbA1c concentrations±SEM (n = 10) in mice immunized (days 0, 14, 28, 42) with Qβ-GIP or Qβ VLPs and fed a high fat diet are shown. No significant difference between vaccinated and control group was observed as determined by two-sided t-test. (E–F) Female mice were immunized (days 0, 14, 28, 42 and 122) with 100 µg of Qβ-GIP or Qβ VLPs and placed on a high fat diet. (E) Blood glucose levels during OGTT on day 142. Average blood glucose levels±SEM (n = 5) at the indicated time points are shown. No significant differences were observed between vaccinated and control animals as determined by two-sided t-test (F) Area under the curve during OGTT on days 31, 58, 93 and 133. Blood glucose levels were measured during OGTT and area under the curve calculated for individual animals. Average AUC±SEM (n = 5) for each group is shown. No significant difference was observed between the vaccinated and control group as determined by two-sided t-test.

Figure 5. Vaccination against GIP does not alter lipid metabolism.

(A–C) Female mice were immunized (days 0, 14, 28, 42) with Qβ-GIP or Qβ VLPs and placed on a high fat diet. (A) Serum lipid profile in mice on day 122. Total, vLDL, LDL and HDL cholesterol was measured. Average values±SEM (n = 10) are shown. No significant differences were observed between the groups (p>0.05). (B) Serum TGL and FFA profiles in mice. The percentage of control animals immunized with Qβ VLPs±SEM (n = 5) at the indicated time point are shown for triglycerides and free fatty acids. No significant differences in TGL or FFA levels were observed between the groups (p>0.05). (C) Postprandial lipid clearance in mice. Female mice were immunized (days 0, 14, 28, 41 and 126) with Qβ-GIP or Qβ VLPs and placed on a high fat diet. OLTT were performed at the indicated time points. The left panel shows the OLTT performed on day 36. AUC for all investigated time points is shown on the right. Average triglyceride levels or AUC±SEM (n = 5) are shown for each group. No significant differences were observed between the groups (p>0.05). All statistical analyses were performed by two sided t-tests.