High-protein diet improves sensitivity to cholecystokinin and shifts the cecal microbiome without altering brain inflammation in diet-induced obesity in rats

Abstract

High-protein diet (HPD) curtails obesity and/or fat mass, but it is unknown whether it reverses neuroinflammation or alters glucose levels, CCK sensitivity, and gut microbiome in rats fed a Western diet (WD)-induced obesity (DIO). Male rats fed a WD (high fat and sugar) for 12 wk were switched to a HPD for 6 wk. Body composition, food intake, meal pattern, sensitivity to intraperitoneal CCK-8S, blood glucose, brain signaling, and cecal microbiota were assessed. When compared with a normal diet, WD increased body weight (9.3%) and fat mass (73.4%). CCK-8S (1.8 or 5.2 nmol/kg) did not alter food intake and meal pattern in DIO rats. Switching to a HPD for 6 wk reduced fat mass (15.7%) with a nonsignificantly reduced body weight gain, normalized blood glucose, and decreased feeding after CCK-8S. DIO rats on the WD or switched to a HPD showed comparable microbial diversity. However, in HPD versus WD rats, there was enrichment of 114 operational taxonomic units (OTUs) and depletion of 188 OTUs. Of those, Akkermansia muciniphila (enriched on a HPD), an unclassified Clostridiales, a member of the RF39 order, and a Phascolarctobacterium were significantly associated with fat mass. The WD increased cytokine expression in the hypothalamus and dorsal medulla that was unchanged by switching to HPD. These data indicate that HPD reduces body fat and restores glucose homeostasis and CCK sensitivity, while not modifying brain inflammation. In addition, expansion of cecal Akkermansia muciniphila correlated to fat mass loss may represent a potential peripheral mechanism of HPD beneficial effects.

Keywords: blood glucose; body composition; gut microbiota; high-fat diet; meal pattern.

Figures

Fig. 1.

Schematic representation of experimental protocols in rats on a Western diet (WD), switched to a high-protein diet (HPD) compared with a normal chow diet (ND). A: feeding behavior monitoring, meal pattern and CCK testing. B: measurements of body composition, blood glucose, cecal microbiota, and brain signals.

Fig. 2.

Body weight and composition of male rats on a WD for 12 wk and either switched to a HPD (WD-HPD) or remaining on a WD compared with controls fed a ND. A–D: six weeks after the HPD switch. E–H: weekly body weight and composition. Data are means ± SE. Rats per group are indicated at the bottom of each bar. *P < 0.05 vs. ND and #P < 0.05 vs. WD (one-way ANOVA or t-test).

Fig. 3.

Weights of adipose tissues in male rats on a WD for 12 wk and either switched to a HPD (WD-HPD) for 6 wk or remaining on WD compared with controls fed a ND. Epididymal (A), mesenteric (B), and brown adipose tissue (C). Data are means ± SE of 8–11 rats/group are indicated at the bottom of each bar in A. *P < 0.05 vs. ND and #P < 0.05 vs. WD (one-way ANOVA).

Fig. 4.

A HPD for 6 wk following a Western diet (WD) for 12 wk improved sensitivity to intraperitoneal CCK-8S compared with rats remaining on WD. A and B: rats maintained on WD. C and D: rats switched to HPD. Data are means ± SE. Rats per group are indicated at the bottom of each bar. *P < 0.05 vs. 0.0 (saline) and #P < 0.05 vs. 1.8 (CCK-8S dose nmol/kg) (one-way ANOVA).

Fig. 5.

HPD for 6 wk following a WD for 12 wk prevents the hyperglycemia induced by a WD. Blood was collected by a tail prick 1–2 h before the dark phase in groups maintained on a ND or a WD, or switched to HPD (WD-HPD). Data are means ± SE, and rats/group are indicated in each bar. *P < 0.05 vs. ND and #P < 0.05 vs. WD (one-way ANOVA).

Fig. 6.

HPD for 6 wk after a WD for 12-wk shifts the cecal microbiome of DIO rats maintained on a WD, with expansion of Akkermansia and reduction of RF39. A: three measures of α diversity are shown for the three dietary groups. P values were calculated using the Mann Whitney U-test. ***P < 0.0001. B: principal coordinates analysis plot colored by dietary group. C: mean taxonomic composition of the three dietary groups is shown at the phylum and genus level. The colors representing the most abundant taxa are shown to the right. D: operational taxonomic units (OTUs) with differential abundance in the HPD group compared with the WD group in DESeq2 models (q < 0.05) are shown, arranged by genus or higher taxonomic level (f = family, o = order) for OTUs that could not be classified at the genus level. Five OTUs that were associated with change in fat mass in multivariate DESeq2 models are colored based on the direction of the association. Circle size represents the significance of the association.

Fig. 7.

HPD for 6 wk after a WD for 12 wk increased the predicted abundance of metagenomic pathways involved in glycan metabolism. A: fold change in predicted abundance between the HPD and WD groups of 56 metagenomics pathways with q < 0.05 in DESeq2 models. The five differential metagenomic pathways categorized in KEGG as related to glycan metabolism are colored in blue. B: contribution of individual taxa to the significance of the five enriched metagenomics pathways related to glycan metabolism was evaluated using FishTaco. The positive and negative effects of taxa on significance are shown separately for taxa that are more abundant in the HPD group (upper bar) and those more abundant in the WD group (lower bar). Taxa could influence significance either by carrying genes in the selected pathway or by being correlated with other taxa carrying the gene. “Other” refers to the total effect of all other taxa not specifically depicted in the graphs.

Fig. 8.

Altered expression of cytokines and mammalian target of rapamycin (mTOR) in the hypothalamus of rats on a WD were not modified by switching to a high-protein diet (WD-HPD) for 6 wk. Data are means ± SE; number of rats/group is indicated at the bottom of the column in the first graph. *P < 0.05 vs. ND (one-way ANOVA).

Fig. 9.

Altered expression of cytokines, POMC, and gut neuropeptide receptors in the dorsal medulla of rats on a WD was not modified by switching to a HPD (WD-HPD) for 6 wk. Data are means ± SE, n = 6–11. *P < 0.05 vs. ND (one-way ANOVA). GLP-1R, glucagon-like peptide 1 receptor; IRa, insulin receptor A; LepR, leptin receptor.