Recombinant Incretin-Secreting Microbe Improves Metabolic Dysfunction in High-Fat Diet Fed Rodents

Abstract

The gut hormone glucagon-like peptide (GLP)-1 and its analogues represent a new generation of anti-diabetic drugs, which have also demonstrated propensity to modulate host lipid metabolism. Despite this, drugs of this nature are currently limited to intramuscular administration routes due to intestinal degradation. The aim of this study was to design a recombinant microbial delivery vector for a GLP-1 analogue and assess the efficacy of the therapeutic in improving host glucose, lipid and cholesterol metabolism in diet induced obese rodents. Diet-induced obese animals received either Lactobacillus paracasei NFBC 338 transformed to express a long-acting analogue of GLP-1 or the isogenic control microbe which solely harbored the pNZ44 plasmid. Short-term GLP-1 microbe intervention in rats reduced serum low-density lipoprotein cholesterol, triglycerides and triglyceride-rich lipoprotein cholesterol substantially. Conversely, extended GLP-1 microbe intervention improved glucose-dependent insulin secretion, glucose metabolism and cholesterol metabolism, compared to the high-fat control group. Interestingly, the microbe significantly attenuated the adiposity associated with the model and altered the serum lipidome, independently of GLP-1 secretion. These data indicate that recombinant incretin-secreting microbes may offer a novel and safe means of managing cholesterol metabolism and diet induced dyslipidaemia, as well as insulin sensitivity in metabolic dysfunction.

Figure 1

Microbial-Produced GLP-1 Displays Insulinotropic Activity Comparable to Synthetic Peptide. (A) pNZ44-KGLP-1 plasmid construct with KGLP-1 amino acid sequence; (B) Overlaid MALDI-TOF spectra of PNZ (red) and GLP1 (blue) spent cell-free growth medium HPLC fractions, with KGLP-1 peak (3512 Da) indicated by arrow. Inset graph displays KGLP-1 production during culture of PNZ (red) and GLP1 (blue) microbes, as assessed by ELISA of cell-free media; (C) Insulin production by Rinm5F-Beta cells treated with Lactobacillus paracasei NFBC 338 PNZ or GLP1 filtrate preparations, or pure KGLP-1 and GLP-1 peptides (50uM). Data was analysed by one-way ANOVA, significant differences are represented by ****(p < 0.0001) and plots depict replicates with mean and SEM.

Figure 2

Short-Term GLP-1-Secreting Lactobacillus paracasei Intervention Modulates Lipid but Not Glucose Metabolism. Experiment I: (A) Schematic of study timeline, with high-fat diet represented by yellow bar and intervention by broken grey bar. (B) Weight gain of PNZ (red) and GLP1 (light blue) rats over treatment period. (C) Oral Glucose Tolerance Test: blood glucose of PNZ (red) and GLP1 (light blue) treatment groups before and after glucose challenge. (D) Fasting serum triglycerides, (E) total cholesterol, (F) apoB-48, (G) LDL-C and (H) HDL-C of PNZ (red) and GLP1 (light blue) treatment groups at cull for experiment I. (I) Triglyceride-rich lipoprotein cholesterol (TLR-C) and (J) triglyceride-rich lipoprotein-mediated atherogenic dyslipidaemia (TLR-AD) index were calculated from the data above. Figures represent the mean with SEM or max and min bars (n = 9). Significant differences are represented by *(p < 0.05).

Figure 3

Lactobacillus paracasei Attenuates Mouse Adiposity Independently of GLP-1-Expression. Experiment II: (A) Mouse trial design and procedures detailed above, with high-fat diet represented by yellow bar and intervention by broken grey bar. (B) Delta change in body weights over the pre-feeding (white background) and intervention (red background) periods for HFC (dark red; n = 13), PNZ (light red; n = 14), GLP1 (light blue; n = 14) and LFC (dark blue; n = 14). Tall dark red boxes represent metabolic test weeks (IPGTT/ITT and mixed meal gavage, respectively). (C) Epididymal (EAT), subcutaneous (SAT) and mesenteric adipose tissue (MAT), and liver tissue weights are also depicted. Data was analysed by one-way ANOVA with Bonferroni correction, significant differences are represented by *(p < 0.05), **(p < 0.01), ***(p < 0.001), ****(p < 0.0001), where the asterisks are coloured according to the group with which a significant difference was recorded. Plots depict replicates with mean and SEM.

Figure 4

GLP-1-Expressing Lactobacillus paracasei Improves Glucose & Cholesterol Metabolism While Promoting Glucose-Dependant Insulin Secretion. Experiment II: HFC (n = 7) and LFC (n = 7) animals were assessed for 1 g/kg glucose (IPGTT) and 0.75 IU/kg insulin (ITT) tolerance at week 12 of feeding, prior to commencement of interventions. (A) IPGTT; (B) IPGTT area under the curve (AUC); (C) ITT; (D) ITT AUC. At week 22 of feeding/week 10 of intervention HFC (n = 7), PNZ (n = 14), GLP1 (n = 13) and LFC (n = 7) animals were again assessed for glucose and insulin tolerance. (E) IPGTT glucose levels; (F) IPGTT glucose AUC; (G) IPGTT insulin levels (T0 and T15); (H) IPGTT insulin AUC; (I) ITT glucose levels; (J) ITT glucose AUC. Animals were assessed for lipid and cholesterol metabolism by oral gavage with a complete meal (K,L) Ensure Plus, Abbott Nutrition). Data was analysed by one-way ANOVA with Bonferroni correction. Significant differences are represented by *(p < 0.05), **(p < 0.01), ***(p < 0.001), ****(p < 0.0001), where the asterisks are coloured according to the group which they represent. Plots depict replicates with mean and SEM.

Figure 5

Lactobacillus paracasei Modulates Host Amino Acid, Biogenic Amine & Phosphotidylcholine Metabolism Independently of GLP-1-Expression. (A) Experiment II: Serum metabolome heatmap, with dendogram clustering according to sample likeness. (B) Principle Coordinate Analysis (PCoA) plot displays HFC (green), PNZ (.blue) and GLP1 (red) samples. (C,D) Quantitative data is displayed for the metabolites which were significantly altered by GLP1 or PNZ; this includes amino acids and biogenic amines, as well as diacyl-phosphatidylcholine (PC aa) and acyl-alkyl-phosphatidylcholine (PC ae). *(q < 0.05), **(q < 0.01), ***(q < 0.001), ****(q < 0.0001) represent significant differences between both PNZ/GLP1 and HFC. Plots depict individual replicates (n = 10) with mean and SEM.