Regulatory mechanisms of fasting

Abstaining from food intake, i.e. fasting, elicits a multi-organ response that maintains the physiological homeostasis of an organism despite a constant requirement for energy. This distinguishes it from the fed state in which the organism maintains homeostasis utilizing ingested nutrients. An intricate network of evolutionarily conserved transcriptional, translational and post-translational regulatory mechanisms underlie the adaptive transition from the fed to the fasted state and provide the basis for the survival of every fasting bout. In recent years, numerous studies documented the salutary health benefits the various fasting protocols elicit in human and animal models. The regulatory mechanisms that facilitate the feeding-fasting transition are incompletely understood. Hence, the goal of this thesis was to characterize a number of transcriptional and translational regulatory mechanisms relevant to fasting. Their uncovering and understanding in the context of basic animal and human physiology is expected to provide new avenues for clinical mitigation of metabolic diseases.

Chapter 1 summarizes the benefits of fasting and describes the feeding-fasting transition. Intermittent fasting (IF), time-restricted feeding (TRF), alternate-day fasting (ADF) and periodic fasting (PF) are the main fasting regimes that are currently being researched. In humans and animal models these regimes normalize HbA1c and glucose levels, improve insulin sensitivity and blood lipid parameters, reduce blood pressure and induce weight loss. In animal models there is evidence for the extension of life span and the mitigation of disease in experimental models for diabetes, cardiovascular disease, neurological disease and cancer.

Key metabolic organs to understand feeding-fasting cycles are the liver, muscle tissue, adipose tissue and the brain. The fed state commences with the ingestion of a meal. Insulin is a key hormone that signals the ‘fed’ state due to its numerous central functions in nutrient uptake and storage. It triggers the uptake of glucose into liver, adipose and muscle tissue, and its utilization to produce ATP via glycolysis. The oxidation of FFA for ATP is generally low in most tissues when Insulin levels are high. In the fed state, the adipose tissue remains the principal tissue for the clearance and storage of FFA in the form of TG inside of lipid droplets. The brain mostly metabolized glucose. The body transitions from the fed to the fasted state when the last meal is fully digested and absorbed. In the first phase, hepatic glycogenolysis and gluconeogenesis are augmented and jointly maintain blood glucose levels. Together with other gluconeogenic substrates like lactate, amino acids are readily liberated upon fasting to maintain blood glucose levels – at the expense of muscle mass. However, uncontrolled catabolism of muscle mass is expected to impede with the long-term survival of the fasting individual. Hence with the extension of the fast from several hours into days, a second vital priority becomes to reduce the reliance on glucose for the generation of ATP. In the second phase plasma amino acid levels and urinary excretion decline, while the body reduces total glucose production. Governed by a dramatic hormonal switch from insulin and leptin to glucagon, cortisol and epinephrine the organism increases rates of β-oxidation for the production of ATP to substitute glucose as a substrate. Activation of adipose tissue lipolysis causes levels of free fatty acids (FFA) and glycerol to increase after 24h of fasting. Additionally, TCA cycle intermediates such as oxaloacetate are utilized to maintain basal levels of gluconeogenesis, resulting in a reduction in hepatic TCA cycle activity. As a result, FFA’s released from the adipose tissue accumulate in the liver and condense into to the ketone bodies acetoacetate (AcAc) and β-hydroxybutyrate (βOHB). Ketone bodies can rise up to 8 mM and, in addition to FFA, are increasingly oxidized by most tissues, including the brain which can derive up to 2/3rd of its energy requirements from ketone body oxidation after several weeks of fasting. The maximal duration of the fast depends on the quantity of endogenous energy. It is estimated that a 70 kg man with 15 kg of adipose tissue could a survive a 70-90 day fast, based on a basal energy expenditure of 1800 kcal/day.

Peroxisome Proliferator-Activated receptor α (PPARα) and cAMP-Responsive Element Binding Protein 3-Like 3 (CREB3L3) are transcription factors governing lipid metabolism in the liver. In Chapter 2 of this thesis, we investigated the interrelationship of the transcription factors PPARα and CREB3L3 in regulating the adaptive response of the liver to feeding-fasting cycles. Male wild-type, PPARα–/–, CREB3L3–/– and combined PPARα/CREB3L3–/– mice were subjected to a 16-hour fast or 4 days of ketogenic diet feeding and whole genome expression analysis was performed on liver samples. The results indicate great plasticity in the degree of synergism and co-dependence between both transcription factors that was dependent on nutritional status. Under conditions of overnight fasting, the effects of PPARα ablation and CREB3L3 ablation on plasma triglyceride, plasma β-hydroxybutyrate, and hepatic gene expression were largely disparate, and showed only limited interdependence. Gene and pathway analysis underscored the importance of CREB3L3 in regulating (apo)lipoprotein metabolism, and of PPARα as master regulator of intracellular lipid metabolism. By contrast, a strong interaction between PPARα and CREB3L3 ablation was observed during ketogenic diet feeding. Loss of CREB3L3 reduced expression of PPARα and its target genes involved in fatty acid oxidation and ketogenesis. In stark contrast, the hepatoproliferative function of PPARα was markedly activated by loss of CREB3L3. Overall, these data indicate that CREB3L3 ablation uncouples the hepatoproliferative and lipid metabolic effects of PPARα. Except for the shared regulation of a very limited number of genes, the roles of PPARα and CREB3L3 in hepatic lipid metabolism are clearly distinct and are highly dependent on dietary status.

Lipoprotein lipase (LPL), a ubiquitously expressed enzyme, is responsible for the hydrolysis of plasma triglycerides on the endothelial lumen to accommodate tissue specific requirements for fatty acids. One protein that increases in abundance during the fasted state and regulates the clearance of triglycerides is Angiopoietin-like 4 (ANGPTL4). In Chapter 3 of this thesis we performed a 24h fasting study to investigate the mechanistic control of adipose tissue LPL activity by ANGPTL4 in humans. Our results indicate that ANGPTL4 levels in human adipose tissue are increased by fasting, likely via increased plasma cortisol and free fatty acids, and decreased plasma insulin, resulting in decreased LPL activity. We conclude that ANGPTL4 levels in human adipose tissue are increased by fasting, resulting in decreased LPL activity. This likely serves to direct TG away from adipose tissue and prevent futile cycling of fatty acids in the fasted state.

In Chapter 4 we examined the mechanism of how ANGPTL4 regulates LPL activity. Previous research indicated that ANGPTL4 regulates LPL activity by cleavage. Treatment of different mouse adipocytes with the PCSK-inhibitor Dec-RVKR-CMK markedly decreased LPL cleavage, indicating that LPL is cleaved by PCSKs. Silencing of Pcsk3/furin significantly decreased LPL cleavage in cell culture medium and lysates of adipocytes. Remarkably, PCSK-mediated cleavage of LPL in adipocytes was diminished by Angptl4 silencing and was decreased in adipocytes and adipose tissue of Angptl4–/– mice. Differences in LPL cleavage between Angptl4–/– and wild-type mice were abrogated by treatment with Dec-RVKR-CMK. Induction of ANGPTL4 in adipose tissue during fasting enhanced PCSK-mediated LPL cleavage, concurrent with decreased LPL activity, in wild-type but not Angptl4–/– mice. In adipocytes levels of N-terminal LPL were markedly higher in wild-type compared to Angptl4–/– adipocytes, suggesting an intracellular mode of action for the stimulation of PCSK-mediated LPL cleavage by ANGPTL4. We conclude that ANGPTL4 promotes PCSK-mediated intracellular cleavage of LPL in adipocytes, likely contributing to regulation of LPL in adipose tissue.

ANGPTL4 impacts plasma TG levels via the regulation of LPL activity. This renders ANGPTL4 an attractive target for correcting dyslipidemia. However, whole body ANGPTL4 inactivation in mice fed a high fat diet causes chylous ascites, an acute-phase response, and mesenteric lymphadenopathy. In Chapter 5, the mechanisms underlying the adverse effect of whole-body inactivation of ANGPTL4 are studied using Angptl4-hypomorphic and Angptl4–/– mice. Angptl4-hypomorphic mice with partial expression of truncated N-terminal ANGPTL4 exhibited reduced fasting plasma triglyceride, cholesterol, and non-esterified fatty acid levels, strongly resembling Angptl4–/– mice. However, during high fat feeding, Angptl4-hypomorphic mice showed markedly delayed and attenuated elevation in plasma serum amyloid A and much milder chylous ascites than Angptl4–/– mice, despite similar abundance of lipid-laden giant cells in mesenteric lymph nodes. Compared with the absence of ANGPTL4, low levels of N-terminal ANGPTL4 mitigate the development of chylous ascites and an acute-phase response in mice. However, since side-effects were still observed, whole body inactivation of ANGPTL4 may not be suitable as a clinical strategy for correcting dyslipidemia.

Intermittent fasting and caloric restriction-type interventions elicit a myriad of health benefits and are expected to provide clinical benefit for the prevention and treatment of obesity, type 2 and potentially cardiovascular disease and cancer. β-hydroxybutyrate (βOHB), levels of which increase to the low mM range during fasting, has been proposed as a potent signaling molecule mediating some of the beneficial effects of fasting interventions. Chapter 6 explores whether the ketone body β-hydroxybutyrate (βOHB) influences differentiation in vitro and regulates genes expression in several primary cell types. We found that βOHB did not impact the differentiation process in C2C12 skeletal muscle cells and 3T3-L1 adipocytes. Transcriptomics analysis revealed overall benign effects of βOHB on gene expression after 6h. βOHB did not consistently regulate genes and pathways in primary adipocytes, macrophages, myotubes and hepatocytes. In contrast, equimolar concentrations of the structurally related HDAC inhibitor butyrate prevented differentiation in C2C12 skeletal muscle cells and 3T3-L1 adipocytes and induced dramatic and consistent gene expression changes in all cell types. Our results suggest that βOHB unlikely mediates the beneficial effects of fasting via regulating gene expression or influencing cellular homeostatic processes such as differentiation. Additionally, in the tested primary cell types, βOHB’s gene expression signature is disparate from the one of butyrate suggesting that βOHB’s HDAC inhibitory capacity may be negligible.

In conclusion, in this thesis we characterized a number of transcriptional and translational regulatory mechanisms relevant to fasting. Depending on nutritional status the hepatic transcription factors PPARα and CREB3L3 show great plasticity in regulating gene expression in the liver (Chapter 2). Furthermore, we were able to confirm the functional importance of human ANGPTL4 for the regulation of LPL activity in adipose tissue during fasting (Chapter 3). The role of ANGPTL4 in regulating LPL activity in non-adipose tissues during fasting is currently unclear. In contrast to mouse adipose tissue, where PCSK3 is involved in the degradation of LPL cleavage products, the precise mechanisms of LPL cleavage in human adipocytes remains to be addressed in future studies (Chapter 4). Current evidence on the potential whole body targeting of ANGPTL4 to correct dyslipidemia is unfavorable yet future studies will have to address the efficacy and safety of tissue-targeted approaches (Chapter 5). Lastly, our results in Chapter 6 in the context of currently available literature do not support the current portrayal of βOHB acting as a pleiotropic signaling molecule. Future research should focus on further defining in which circumstances and via which mechanisms βOHB displays robust signaling effects.