Molecular and physiological assessment of metabolic health : adipose tissue, transcriptome analysis and challenge tests

Research output: Thesisinternal PhD, WU


Summary of main findings

Maintenance of metabolic health not only ensures that energy is made available in times of need and stored in times of excess, but also prevents resistance to nutritional cues, ectopic lipid accumulation and dysfunction of metabolic organs. The proportion of humans that is at risk for reduced metabolic health increases worldwide due to the current epidemic of obesity and the increase in both mean and maximum life span. Better understanding of the various factors that influence metabolic health may offer opportunities to fight this threat to human health. This thesis aims to assess metabolic health using transcriptome analysis and non-invasive challenge tests. Special focus is on the development and validation of InCa-based non-invasive challenge tests. In most chapters of this thesis white adipose tissue (WAT) formed the major organ of interest because of its key role in whole-body energy homeostasis. WAT function was, among others, studied with whole-genome gene expression analysis, which, compared to single parameter analysis, extends the scale and depth of understanding biological processes.

Metabolic health was also quantified as metabolic flexibility, with the use of non-invasive, indirect calorimetry (InCa) based challenge tests. One of the InCa based challenge tests described in this thesis, the oxygen restriction (OxR) challenge, is a novel approach to investigate metabolic flexibility in mice. In each study, OxR was applied acute ([O2] reduction within 30 minutes) and for a short period of 6 hours in fasted mice. The other two InCa-based challenge tests: fasting and re-feeding and fasting and glucose consumption are nutrient-based and were described previously, although in different formats and settings.


In chapter 2 we demonstrate that dietary restriction on a high-fat diet (HF-DR) improves metabolic health of mice compared with mice receiving the same diet on an ad libitum basis (HF-AL). Already after five weeks of restriction, the serum levels of cholesterol and leptin were significantly decreased in HF-DR mice, whereas their glucose tolerance and serum adiponectin levels were increased. The body weight and measured serum parameters remained stable in the following 7 weeks of restriction, implying metabolic adaptation. To understand the molecular events associated with this adaptation, we analysed gene expression in WAT with whole genome microarrays. HF-DR strongly influenced gene expression in WAT; in total, 8643 genes were differentially expressed between both groups of mice, with a major role for genes involved in lipid metabolism and mitochondrial functioning. DR also increases mitochondrial density in WAT. These results show that WAT, indeed, has an important role in the improvement of metabolic health of dietary restricted mice and suggest that the development of substrate efficiency plays an important role in the observed changes in health status. Finally, mitochondrial density might be used as a marker for WAT health status.


Chapter 3 shows how indirect calorimetry can be used to noninvasively assess metabolic and age-related flexibility in mice. In this study, we tested the sensitivity and response stability over time of three InCa-based treatments in old versus adult mice. For the first treatment, diurnal patterns of respiratory exchange ratio were followed for 24 hours under standard conditions. For the second and third treatment, which were both based on a challenge approach, mice were fasted and either received a glucose bolus to test switch-effectiveness from fat to glucose oxidation (Treatment 2), or were exposed to oxygen restriction (OxR, Treatment 3) in the InCa system, which was introduced as a novel approach to asses metabolic flexibility. Opposite to the mice that were dietary restricted (chapter 2), aging appeared to increase adiposity and decrease WAT mitochondrial density, which further suggests that WAT mitochondrial density might be used as a marker for WAT health. We observed that the test results of the first treatment were not stable between test periods, possibly because of behavioural differences within the group of old mice between both measurements. For the second treatment, no differences between groups were observed. With Treatment 3, however, stable significant differences could be detected: old mice did not maintain reduced oxygen consumption under OxR during both measurements, whereas adult mice did. Further biochemical and gene expression analyses showed that OxR affected glucose and lactate homeostasis in liver and WAT of adult mice, supporting the observed differences in oxygen consumption. This was the first study to show that InCa analysis of the response to OxR is a sensitive and reproducible treatment to noninvasively measure age-impaired metabolic health in mice. Evaluation of metabolic health under non-challenged conditions may be confounded by behavioural-induced variation between animals


The study described in chapter 4 followed up on the promising results that were obtained with the OxR challenge in chapter 3. In this study we tested whether OxR can also be used to reveal diet-induced health effects in an early stage. Early detection of diet-induced health effects might shorten animal experiments and reduce costs and age-related variation. Timely identification may increase options for reversal. Mice were exposed to a low-fat (LF) or high-fat (HF) diet for only 5 days, after which they were exposed to OxR or remained under normoxic conditions. The response to OxR was assessed by calorimetric measurements, followed by analysis of gene expression in liver and WAT. A novelty described in this chapter was the analysis of serum markers for protein glycation and oxidation, to detect differences in the response to OxR between LF and HF mice. Although HF feeding increased body weight, HF and LF mice did not differ in indirect calorimetric values under normoxic conditions and in a fasting state. Exposure to OxR however, increased oxygen consumption and lipid oxidation in HF mice versus LF mice. Furthermore, OxR induced gluconeogenesis and an antioxidant response in the liver of HF mice, whereas it induced de novo lipogenesis and an antioxidant response in eWAT of LF mice, indicating that HF and LF mice differed in their adaptation to OxR. OxR also increased serum markers of protein glycation and oxidation in HF mice, whereas these changes were absent in LF mice. From this study we concluded that OxR is a promising new method to test food products on potential beneficial effects for metabolic health.


The study described in chapter 5 aimed to assess differences in metabolic health of mice on iso-caloric diets differing in fatty acid composition using the OxR challenge. We also implemented a fasting and re-feeding challenge. One diet, the HFpu diet, predominantly contained poly-unsaturated fatty acids (PUFAs), which are considered to be healthier than saturated fatty acids (SFAs) that mainly made up the fat component of the second diet, the HFs diet. Since health effects of fatty acids also depend on the ratio of dietary omega-6 to omega-3 PUFAs (n6/n3 ratio), this ratio was kept similar between both diets. Mice received the isocaloric high-fat diets for six months, during and after which several biomarkers for health were measured. We found that HFpu and HFs diets only induce minor differences in static health markers: HFpu and HFs mice did not differ in body weight, total adiposity, adipose tissue health, serum adipokines, whole body energy balance, or circadian rhythm. HFpu and HFs mice also had a similar glucose tolerance, even though HFs mice had more triglycerides in liver and skeletal muscle and larger adipocytes in the eWAT depot. Interestingly, HFs mice were less flexible in their response to both fasting and re-feeding and OxR, which shows the relevance and sensitivity of InCa-based challenge tests. We concluded that InCa-based challenge tests are a valuable contribution to the analysis of metabolic health in mice. Challenge tests in the InCa system may, furthermore, reveal relevant consequences of small changes in metabolic health status, such as adipocyte hypertrophy or ectopic lipid storage.


Chapter 6 describes an in-depth study to the response to OxR both at whole body level using InCa and serum metabolomics (amino acids and (acyl)carnitines) and at WAT level using transcriptomics and the analysis of amino acid and (acyl)carnitine levels. Serum and tissue amino acids levels indicate the level of protein catabolism and certain amino acids are, typically, increased in obese individuals. Serum and tissue (acyl)carnitine levels indicate the rate and completeness of mitochondrial fatty acid oxidation; serum acylcarnitine levels are significantly increased in individuals that suffer from ambient oxygen restriction. The metabolic adaptation to OxR was studied in diet-induced moderately obese mice that received a high-fat diet (HFpu diet, as in chapter 5) for 6 weeks, which is expected to lead to WAT expansion and possibly to reduce oxygen availability in WAT. We found that OxR reduced mitochondrial oxidation at whole-body level, as shown by a reduction in whole-body oxygen consumption and an increase in serum long-chain acylcarnitine levels. WAT did not seem to contribute to this serum profile, since only short-chain acylcarnitines were increased in WAT and gene expression analysis indicated an increase in mitochondrial oxidation, based on coordinate down-regulation of Sirt4, Gpam and Chchd3/Minos3. In addition, OxR did not induce oxidative stress in WAT, but increased molecular pathways involved in cell growth and proliferation. OxR increased levels of tyrosine, lysine and ornithine in serum and of leucine/isoleucine in WAT. This study shows that OxR limits oxidative phosphorylation at whole-body level, but in WAT compensatory mechanisms seem to operate. The down-regulation of the mitochondria-related genes Sirt4, Gpam, and Chchd3 may be considered as a biomarker profile for WAT mitochondrial reprogramming in response to acute exposure to limited oxygen availability.


To conclude, the work presented in this thesis provides more insight in the analysis of metabolic health in mice with the use of transcriptome analysis and InCa-based challenge tests. We show that non-invasive tests using the InCa-system are more likely to reveal differences in metabolic flexibility than invasive challenge tests, such as the oral glucose tolerance test. Furthermore, we show that the challenge approach is more sensitive than analysis of metabolic health under non-challenged (free-feeding) conditions. Transcriptome analysis proved to be very valuable to provide in-depth molecular understanding of the mechanisms underlying reduced or improved metabolic health. Ideally, transciptomic or metabolomic approaches should be integrated with InCa-based challenge tests to further extent physiological understanding of diet-induced health effects.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
  • Keijer, Jaap, Promotor
  • van Schothorst, Evert, Co-promotor
Award date26 May 2015
Place of PublicationWageningen
Print ISBNs9789462573017
Publication statusPublished - 2015


  • mice
  • metabolism
  • health
  • adipose tissue
  • transcriptomes
  • metabolic disorders
  • physiology
  • laboratory animals

Fingerprint Dive into the research topics of 'Molecular and physiological assessment of metabolic health : adipose tissue, transcriptome analysis and challenge tests'. Together they form a unique fingerprint.

  • Projects

    Cite this