Recovery from critical illness: role of autophagy and epigenetics.

The development of critical care medicine over the last 50 years allowed patients to survive acute life-threatening conditions. Whereas most of these patients recover within a few days, a substantial number of critically ill patients develop multiple organ failure (MOF), which prolongs the dependency on intensive care and increases health care costs. Despite major advances in intensive medical care, MOF is still a leading cause of morbidity and mortality. The pathogenesis of MOF is complex and incompletely understood. Recent studies have suggested a pivotal role for mitochondrial dysfunction and endoplasmic reticulum (ER) stress in the development of MOF. As such, identifying underlying molecular pathways and pharmacological interventions that can decrease these cellular stresses during critical illness hold great promise to improve outcome of the patients. 

Critical illness is also characterized by complex endocrine and metabolic alterations, such as the development of hyperglycemia and hypoaminoacidemia, which are both associated with a poor outcome. The underlying mechanisms of these alterations are incompletely understood. Interestingly, several of the endocrine and metabolic alterations that hallmark critical illness have been shown to influence cellular stress, cellular defense mechanisms and the outcome of critically ill patients. As such, gaining more insight in the mechanisms underlying these pathways may identify new therapeutic targets for the critically ill patient.

The general objective of this PhD project was to gain more insight in metabolic mechanisms of organ failure during critical illness. More specifically, we studied the role of autophagy, a crucial cellular defense mechanism, in the pathogenesis of critical illness-induced organ failure, as well as the role of FGF21 and glucagon in mediating cellular stress and metabolic alterations during critical illness.

In a first part, we investigated the role of hepatic autophagy in safeguarding liver function during critical illness in a validated mouse model, by genetically manipulating the hepatic autophagy pathway. Hepatic autophagy deficiency during critical illness aggravated mitochondrial dysfunction and suppressed the adaptive unfolded protein response (UPR), a crucial defense mechanism against ER stress. This resulted in an aggravation of critical illness-induced liver damage, demonstrating that autophagy is an important metabolic pathway to prevent and resolve critical illness-induced organ failure. In another mouse study, we investigated whether the clinically available pharmacological autophagy activators promethazine and carbamazepine are able to stimulate autophagy during critical illness. Unfortunately, administration of these compounds did not stimulate hepatic autophagy and did not prevent liver damage during critical illness.

In a second part, we investigated whether fibroblast growth factor 21 (FGF21), a newly identified hormone, is elevated and can be manipulated by metabolic interventions during critical illness and assessed its role during critical illness. Plasma FGF21 concentrations were strongly induced in critically ill mice and human patients and correlated with the severity of disease. The rise in FGF21 was most likely due to an increase in hepatic fgf21 expression, as was shown in a rabbit model of critical illness.  This increased fgf21 expression correlated strongly with mitochondrial dysfunction and markers of the integrated stress response, the latter being a branch of the UPR. Subsequently, the effect on FGF21 of tight glycemic control, as a therapy that has been shown to decrease cellular stress during critical illness, was studied in a human and a rabbit study. Tight glycemic control decreased serum FGF21 concentrations in critically ill human patients, which statistically explained at least part of the mortality benefit of this intervention. In the rabbit study, tight glycemic control decreased hepatic fgf21 expression, which strongly correlated with attenuation of mitochondrial dysfunction and of the integrated stress response. In a subsequent mouse study, we investigated the role of FGF21 during critical illness, by genetically interfering with FGF21 availability during critical illness. FGF21 deficiency only moderately decreased glycemia and did not affect other investigated metabolic and endocrine pathways. FGF21 deficiency during critical illness, however, inhibited hepatic autophagy and increased the integrated stress response, demonstrating that FGF21 inhibits cellular stress during critical illness possibly by promoting autophagy. Altogether, these results indicated that FGF21 is an adaptive cellular stress-induced hormone during critical illness.

In the last part, we investigated whether glucagon is elevated during critical illness and whether it can be modulated by parenteral nutrition (PN), and we assessed its role in modulating metabolic and cellular stress. Glucagon was elevated in critically ill mice and human patients and correlated with the severity of disease. Providing glucose and insulin to critically ill patients did not decrease glucagon, unlike in healthy humans. In contrast, providing PN containing amino acids to critically ill patients strongly increased plasma glucagon. This was further investigated in a validated mouse model of critical illness, showing that the amino acid component of PN is responsible for the increase in plasma glucagon concentrations. In the same mouse model, we investigated the role of glucagon during critical illness by immunoneutralizing it during critical illness. Glucagon immunoneutralization only transiently affected FGF21, glucose and lipid metabolism during critical illness, whereas it profoundly suppressed hepatic amino acid catabolism, reversing the critical illness-induced hypoaminoacidemia. Immunoneutralizing glucagon did not affect muscle wasting. Furthermore, providing amino acids, with the intention to counteract critical illness-induced muscle wasting, only increased glucagon availability and aggravated the hepatic amino acid catabolism, without affecting muscle wasting during critical illness. Also in the same mouse model, we showed that elevated glucagon concentrations did not stimulate hepatic autophagy, which was already suppressed from early onwards in the course of critical illness.

In conclusion, we demonstrated that insufficient activation of autophagy is an important contributor to critical illness-induced liver damage/failure, by increasing mitochondrial dysfunction and ER stress. Next, we identified FGF21 as a new adaptive cellular stress-induced hormone during critical illness, which promotes hepatic autophagy and decreases hepatic cellular stress. Finally, we showed that glucagon is a key mediator of the derangements in amino acid metabolism during critical illness, by promoting hepatic amino acid catabolism, and that early initiation of parenteral nutrition and especially amino acids during critical illness increases amino acid catabolism by stimulating glucagon. These data increase our understanding of the pathophysiology of critical illness and open perspectives for new therapies for treating critically ill patients.