Extended indirect calorimetry as a physiological phenotyping tool in mouse nutritional intervention studies, with a focus on metabolic programming by starches

Indirect calorimetry (InCa) is an essential tool for human and animal studies of energy metabolism. InCa measures the metabolic gases consumed (oxygen, O2) and produced (carbon dioxide, CO2) by the organism to calculate energy expenditure (EE). The ratio of CO2 produced to O2 consumed, or the respiratory exchange ratio (RER), indicates substrate utilisation (carbohydrate, fat, and protein) at the whole body level. Significant improvements in commercial InCa systems for rodents have increased its accuracy and resolution, yet there have been few attempts to extend the technique by measuring other physiological gases. These gases can be of microbial origin, like hydrogen (H2) and methane (CH4), which result from fermentation by the gut microbiome. Furthermore, conventional InCa cannot distinguish between the oxidation of exogenous (e.g. dietary) and endogenous (stored in the body) substrates. Stable isotopic tracers, like 13C-enriched nutrients, make this distinction possible by analysis of the ratio of 13CO2/12CO2 in expired air. These gases (H2, CH4, 13CO2, 12CO2) are rarely measured in mouse studies and, when they are, they need expensive equipment and considerable time and effort.

Dietary starches can be lowly or highly digestible. Lowly digestible starches generally produce lower glycaemic responses, provide fermentative substrates for the intestinal microbiota, and are thought to prevent excessive adiposity and favour metabolic health. There are some indications that starch digestibility has different metabolic effects in females and males. The evidence for this is very limited and, where both sexes have been studied, little attention has been paid to other aspects of carbohydrate metabolism, besides oral glucose tolerance tests and static tests for glycaemia. Moreover, despite the potential of starches to impact metabolism, there are no evidence-based recommendations for starch intake for young children. This is important because early life nutrition has the power to condition the metabolic function of the organism in the long term, a phenomenon known as metabolic programming. It is not known whether starch digestibility can program the organism metabolically, neither whether programming by starches can occur in the post-weaning period, a period where the organism faces a dramatic change in dietary macronutrient composition and density.

This thesis aimed to 1) develop an show the added value of an extended mouse metabolic phenotyping tool based on InCa for the real-time study of microbiota activity and the oxidation of exogenous vs endogenous substrates; and 2) to apply this tool to study the direct and metabolic programming effects of starches consumed during the early post-weaning period.

In Chapter 2, we aimed to examine whether we would be able to study microbiota activity non-invasively, continuously and in real time by extending a commercial InCa system (eInCa) with H2 and CH4 sensors (eInCa). Hydrogen production was circadian and depended on food intake and starch digestibility, as tested in mice fed a lowly (LDD) or a highly digestible-starch diet (HDD). Hydrogen production explained ~20% of the variation in faecal bacterial composition, and correlated with specific bacterial genera known to produce H2 in vitro. No CH4 production could be measured, and this was consistent with the absence of faecal methanogenic archaea. We concluded that eInCa is a useful tool to study diet-microbiota-host interactions in real time.

We then incorporated 13CO2 and 12CO2 sensors into the same eInCa system and aimed to demonstrate their added value in Chapter 3. The system detected to differences in 13CO2 enrichment based on the natural 13C enrichment of the diet and daily food intake patterns. By combining 13CO2 enrichment and conventional InCa data, we were able to quantify the oxidation rates of 13C glucose or 13C palmitate ingested by lean and obese mice, separately from total (exogenous + endogenous) glucose and fat oxidation rates. The oxidation of ingested palmitate was negatively correlated to the animal’s fat mas and positively correlated to metabolic flexibility. We concluded that enabling 13CO2 enrichment analysis in eInCa makes it a powerful tool for the quantification of specific substrate oxidation in physiological and pathophysiological conditions .

In Chapter 4, we aimed to test whether the capacity to oxidise the starch molecule is influenced by prior exposure to an LDD or an HDD, and if this effect is similar for both females and males. The oxidation of a highly digestible starch bolus was higher in mice that consumed LDD for 3 weeks prior. This effect was larger in females than in males, as shown by the faster starch oxidation kinetics of LDD vs HDD females during the early postprandial period, not present in males. However, LDD males exhibited a constantly higher RER, both in fasting conditions and during the postprandial period after consumption of the starch bolus. Small intestinal amylase levels did not explain the higher starch oxidation of LDD vs HDD females. From these data we concluded that short-term consumption of lowly digestible starch enhances the oxidation of the starch molecule, especially in females.

In Chapter 5, we aimed to retest some of the direct effects of starch digestibility and to test whether LDD vs HDD consumed in early post-weaning can programme metabolism in the long-term in both sexes. Direct exposure to LDD led to smaller adipocyte sizes and lower inflammation in the gonadal adipose compartment of females, and decreased body fat mass in males. In both sexes, digestible energy intake and H2 production were directly increased by LDD. Adult females on a HFD that were exposed to LDD during post-weaning had a better metabolic flexibility and lower macrophage infiltration in the gonadal adipose depot. In males, no programming effects on metabolic flexibility and other metabolic outcomes like body composition and fasting glucose, insulin, and adipokine levels were observed. We concluded that the metabolic programming effects of starches consumed in the post-weaning period are subtle and sexually dimorphic.

I discuss the findings of my thesis and provide implications and suggestions for further research in Chapter 6.

To conclude, we have shown the added value of integrating new gas sensors into a commercial InCa system (Chapters 2 and 3). This technological extension yields real-time, continuous, and automated data that can inform microbiology and nutritional studies and beyond. By providing additional information with minimal discomfort to the animal, eInCa can contribute to the replacement, reduction, and refinement (3 Rs) of animal experimentation.