Immunometabolism

mTOR is a serine/threonine protein kinase, which is found in 2 complexes in cells: mTOR complex 1 and 2 (mTORC1 and mTORC2). mTORC1 is activated through the T cell receptor (TCR) and the costimulatory molecule cluster of differentiation 28 (CD28) engagement. However, it can also be activated by growth factors like IL-7 or IL-2 and by metabolites like glucose or amino acids (leucin, arginine or glutamine).[7][6] In contrast, there are more gaps as to how mTORC2 pathway functions, but its activation is also achieved through growth factors as exemplified by IL-2.[6]

When activated mTORC1 negatively regulates autophagy (through inhibiting the ULK complex) and shifts the cell towards aerobic glycolysis, glutaminolysis (through activation of c-Myc) and promotes lipid synthesis and mitochondrial remodelling.[7][6] mTORC2 enhances glycolysis as well, but in contrast to mTORC1, it activates akt, which in turn promotes glucose transporter 1 (GLUT1) membrane deposition. It also further promotes, through other kinases, cell proliferation and survival.[6]

PI3K mediates the phosphorylation of phosphatidylinositol-(4,5)-bisphosphate (PIP2) into phosphatidylinositol-(3,4,5)-trisphosphate (PIP3). PIP3 then serves as a scaffold for other proteins, which contain a pleckstrin homology (PH) domain. It can be activated, just like mTOR, through TCR, CD28 and, unlike mTOR, through another costimulatory molecule: Inducible T-cell COStimulator (ICOS).[6]

The present of PIP3 on a membrane recruits many proteins including phosphoinositide-dependent protein kinase 1 (PDK1), which after its phosphorylation together with mTORC2 activates akt, a serine/threonine kinase. As a result akt promotes GLUT1 membrane deposition and akt also inhibits transcription factor forkhead box O (FoxO), whose inactivation acts in synergy with the mTORC2 above mentioned changes.[6][8]

Both LBK1 and AMPK are serine/threonine kinases acting predominantly opposingly to the aforementioned molecules. From the two, LBK1's activation is less understood, as it is mainly dependants on cellular localization and on many posttranslational modifications. For instance the above-mentioned akt can stimulate LBK1 inhibition through promoting nuclear retention. When activated, LBK1 can activate, apart from other targets, AMPK, whose activation leads to mTORC1 destabilization.[6] Furthermore, it activates ULK complex, phosphorylates p53 and acetyl-CoA carboxylase (ACC), which promotes autophagy, cell cycle arrest and fatty acids oxidation respectively. Since AMPK can also be activated through adenosine monophosphate (AMP) or by glucose insufficiency, it acts as a sensor of starvation and therefore activates many already mentioned catabolic processes, which is in direct contrast with mTOR, which activates myriad of anabolic processes.[6][7][9]

Naive T cells have to be kept in a permanent state of quiescence, until they encounter their cognate antigen. The quiescence state is sustained by tonic TCR signalling and by IL-7. Tonic TCR signalling is necessary to keep the FoxO transcription factor active, which in turn allows for IL-7R transcription. This enables the T cell to survive and proliferate at a low rate. However, during this tonic TCR signalling proteins, that control metabolism, have to be strictly regulated, because their activation could lead to spontaneous exit of quiescence and differentiation into various T cells subset, as exemplified by the uncontrolled activation of PI3K which causes the development of Th1 or Th2.[11]

Both of the aforementioned signals should lead to the mTOR and akt activation, but in quiescence T cells there are tuberous sclerosis complex (TSC) and phosphatase and tensin homolog (PTEN) acting against their activations. Therefore, a naive T cell dependent predominantly on oxidative phosphorylation and has much lower glucose uptake and ATP production than their activated counterparts (effector T cells).[11][7][6]

Quiescence exit begins when a T cell encounters its cognate antigen usually during an infection. The TCR signal together with the costimulation signal lead to downregulation of PTEN and TSC.[11] This causes the phosphorylation cascades of mTOR and akt and many more kinases to be fully activated. These cascades activities result in glucose and glutamine uptake coupled with higher glycolysis and glutaminolysis, which not only supports rapid cell growth, but also further promotes mTOR activation. Furthermore, mTOR stimulates lipid synthesis and mitochondria remodelling, exemplified by increased expression of sterol regulatory element-binding protein (SREBP) and mitochondria undergoing fission, which causes them to function predominantly as biosynthetic hubs, rather than energy production hubs. After their activation and the metabolic reprogramming, T cells compete with one another and consequently, it is very likely that during its effector phase T cells reach a point, where they suffer from lack of nutrients. In such cases AMPK is activated to balance the mTOR signalling and to prevent apoptosis.[11][6][4]

The described scheme of quiescence exit holds true for inflammatory T cells subsets like Th1, Th2, Th17 and cytotoxic T cells. However, mTOR activity can be detrimental when we focus on Tregs. This is shown by the fact that in Tregs high activation of mTORC1 coupled with a higher level of glycolysis leads to the failure of Treg lineage commitment. Therefore, in contrast to inflammatory cell subsets, Tregs rely on oxidative phosphorylation fuelled by lipid oxidation.[11][4] Although, it is important to note that complete suppression of glycolysis leads to enolase (a glycolytic enzyme) binding to a splice variant of Foxp3, which effectively compromises peripheral Tregs abilities to act as immunosuppressive cells.[7][4]

After the infection is cleared most of the activated T cells succumb to apoptosis. However, few of them survive and develop into the memory T cell subsets. For this development the engagement of costimulatory molecules, like CD28, appears to be crucial, as the co-stimulation manifests in mitochondrial morphology, thus allowing for higher oxidative phosphorylation but also retaining the potential to quickly revert to glycolysis.[12][13] Moreover, T cell activation causes an overall increase in acetyl-CoA, which is a substrate for the histone acetylation. As a results, many genes are acetylated and therefore accessible to transcription even after the differentiation into memory subsets, hence allowing memory T cells to rapidly re-express some effector related genes.[12] The aforementioned changes allow T cells to become memory cells, but what exactly drives the memory cell differentiation is still under debate, even though IL-15 seems to be necessary for the T cell memory induction. Recently, asymmetric division of mTORC1, during the first divisions after TCR activation, has been shown to drive the memory cell differentiation in those cells which receive lower amount of mTORC1.[12][13]

Immunometabolism of macrophages is mostly studied in the two opposing populations of macrophages: M1 and M2. M1 macrophages are a pro-inflammatory population induced by LPS or IFNγ. This activation leads, as in the case of T cells, to increase in glycose uptake and glycolysis. What is strikingly different is the Krebs cycle, as in the case of M1 macrophages the cycle is broken at two places. The first break is the conversion of iso-citrate to α-ketoglutarate owing to the downregulation of isocitrate dehydrogenase. Accumulated citrate is subsequently used for lipid and itaconate synthesis, which are both indispensable for M1 macrophages function. The second break at the succinate to fumarate transition occurs probably due to the itaconate production and causes a build up of succinate. This triggers ROS production, which stabilizes HIF-1α. This transcription factor further promotes glycolysis and it is essential for activation of inflammatory macrophages.[10][4]

M2 macrophages are anti-inflammatory cells which need for their induction IL-4. M2 macrophages metabolism is markedly distinct from M1 macrophages due to their unbroken Krebs cycle, which after their activation is fuelled by upragulated glycolysis, glutaminolysis and fatty acid oxidation.[10][4] How the fully operational Krebs cycle exactly translates to M2 macrophages functions is still poorly understood, but the upregulated pathways allow for production of intermediates (mainly acetyl-CoA and S-adenosyl methionine), which are needed for histone modifications of genes targeted by IL-4 signalling.[10]