Autophagy is known to play a considerable role in liver physiology and pathology [17–19]. Zonated regulation of this process may offer not only the possibility to differently connect autophagy with anabolic and catabolic pathways which are usually inversely zonated, but also to influence these pathways in different ways. Since our hypothesis includes both, metabolic regulation through amino acids and morphogen signalling controlling the proportion of zonated functions, the implications for liver metabolism and pathology are very versatile. Some examples are discussed below.
Under well nourished conditions, amino acids (including glutamine and EAA) entering through afferent vessels are high. The suggested regulatory mechanism for periportal autophagy implies that part of the glutamine taken up is re-exported for exchange of leucine which subsequently inhibits autophagy by activating mTORC1. This may favour maintenance of mitochondria for optimally driving urea synthesis and keeping nitrogen balanced. Simultaneously, pericentral FOXO-mediated autophagy may act largely unaffected ensuring protection against increased risk of cell deterioration due to decreasing pericentral oxygen concentrations. However, if such a well nourished condition continues over time, reduced periportal autophagy may raise p62 levels compromising degradation of ubiquitine-proteasome pathway substrates  and eventually resulting in liver pathology.
During starvation, the opposite scenario is likely. Levels of glutamine and EAA in portal blood are quite low. Thus, little leucine may enter the periportal hepatocytes, mTORC1 remains inhibited and autophagy is activated. This mechanism may contribute to the well-known fact that the liver can maintain blood glucose and amino acid levels by sacrificing up to 40% of its protein in an early stage of starvation [7, 21, 22]. This process may include both, periportal and pericentral hepatocytes, since glutamine production in pericentral hepatocytes is increased due to enhanced ammonia levels. Consequently, FOXO-mediated autophagy should also be stimulated during starvation. Interestingly, repeated starvation may cause extension of the GS-positive zone  and, thus, may shift the balance between the two regulatory mechanisms of autophagy in favour of FOXO-mediated autophagy.
Another important issue affected by our hypothesis concerns liver lipid metabolism. Autophagy has recently been found to play an important role in lipid metabolism particularly in liver, because activation may lead to enhanced lipid degradation (also known as lipophagy ), while inhibition may result in a steatotic phenotype . However, the situation appears much more complex. For instance, lipophagy during starvation may have a protecting function  by limiting the puzzling accumulation of triglycerides occurring during a 24 h fasting period  due to flooding the liver with free fatty acids liberated from adipose tissue. Different contributions of periportal and pericentral autophagy may explain the observed focal rather than global distribution of lipid droplets. Furthermore, independent regulation of pericentral autophagy as hypothesized herein offers the possibility for independent regulation of peroxisomal β-oxidation of fatty acids by FOXO-mediated autophagy, because peroxisomes are preferentially found in the pericentral zone. Indeed, treating fasted rats with antilipolytic drugs resulted in changes in peroxisomal rather than mitochondrial enzyme activities . Notably, peroxisome distribution can be enlarged by dihydroepiandrosterone, a drug also enlarging the GS-positive zone  and, thus, the zone of FOXO-mediated autophagy.
The proposed dependence of the regulation of autophagy on Wnt and Hh signalling is of particular interest, since both morphogen signalling pathways can be considered as master regulators of liver zonation. This has been demonstrated for Wnt signalling which controls amino acid, ammonia and carbohydrate metabolism [3, 5, 29] and, via FOXO3 and glutamine synthesis, FOXO-mediated autophagy. The contribution of Hh signalling to the control of liver zonation is still hypothetical in spite of supportive data (for review see ). As shown in other organs, however, Hh signalling controls lipid metabolism in adipose tissue  and autophagy in vascular smooth muscle cells . We assume similar effects to occur in liver, particularly in the periportal zone. Further evidence suggests that autophagy may regulate Wnt signalling by promoting Dishevelled degradation . Taken together, these findings may imply that autophagy is not only subject to regulation by morphogens, but conversely may contribute to shaping graded morphogen action, an as yet unsolved problem in liver . Given the fact that liver zonation seems to be of considerable importance for the development of distinct phenotypic classes of hepatocellular tumors [33, 34], zonated regulation of autophagy may have more impact on the development of liver cancer than thought before .
Moreover, since GS is heterogeneously expressed in many tissues (prominent examples are kidney , skin , and small intestine ) matching inverse gradients of Wnt and hedgehog signalling (for small intestine see ), the dual glutamine-dependent opposing mechanisms described herein, may represent a more general principle for balancing bulk protein turnover by autophagy.