Oncogenic signalling and metabolic alterations are interrelated in cancer cells. mTOR, which is frequently activated in cancer, controls cell growth and metabolism. mTOR signalling regulates amino acid, glucose, nucleotide, fatty acid and lipid metabolism. Conversely, metabolic inputs, such as amino acids, activate mTOR. In this Review, we discuss how mTOR signalling rewires cancer cell metabolism and delineate how changes in metabolism, in turn, sustain mTOR signalling and tumorigenicity. Several drugs are being developed to perturb cancer cell metabolism. However, their efficacy as stand-alone therapies, similar to mTOR inhibitors, is limited. Here, we discuss how the interdependence of mTOR signalling and metabolism can be exploited for cancer therapy. rapamycin treatment. Activation of mTORC1 by growth factors and nutrientsThe activation of mTORC1 is dependent on nutrients and growth factors. In response to nutrients, mTORC1 translocates from the cytoplasm to the lysosomal surface, where it is activated by growth factors via PI3K-AKT signalling. Growth factors, for instance, insulin, activate AKT4 via a cognate receptor, phosphoinositide-dependent kinase 1 (PDK1) and PI3K (Fig. 1). AKT inhibits the TSC1-TSC2 complex5, which is a GTPase-activating protein (GAP) for the small GTPase RHEB6. GTP-bound RHEB directly binds and activates mTORC1 at the lysosome 7,8 (Fig. 1). Nutrient-induced lysosomal translocation of mTORC1Nutrients, in particular amino acids, promote lysosomal localization of mTORC1 via the RAS-related GTP-binding proteins (RAGs) 9 , thereby enabling mTORC1 to encounter RHEB. RAGs are small GTPases that form obligate heterodimers. RAGA or RAGB associates with RAGC or RAGD. In the active state, GTP-bound RAGA or RAGB and GDP-bound RAGC or RAGD bind RAPTOR and thereby recruit mTORC1 to the lysosomal surface. The nucleotide binding status of the RAGs is tightly regulated by amino acids 9 obtained from intracellular synthesis, protein turnover or extracellular sources via specific transporters (Fig. 1; for details, see following sections). Among the amino acids, leucine, arginine and glutamine are the most effective activators of mTORC1. Leucine and arginine bind to sestrin 2 and CASTOR1, respectively, ultimately to activate the RAGs and mTORC1 (for details, see Fig. 1, ref. 10 and references therein). The lysosomal amino acid transporter SLC38A9 promotes mTORC1 activation by exporting essential amino acids to the cytoplasm 11 , where, for example, leucine can bind sestrin 2. Leucine export is stimulated by arginine binding to SLC38A9 (ref. 11 ). Glutamine activates RAGs by promoting glutaminolysis. During glutaminolysis, glutaminase (GLS) and glutamate dehydrogenase (GDH) convert glutamine to αketoglutarate (αKG), which ultimately activates mTORC1 via prolyl hydroxylases (PHDs) by promoting GTP loading of RAGB12 (Fig. 1). Leucine also stimulates αKG production by directly binding and allosterically activating GDH. Furthermore, glutamine activates mTORC1 independently of RAGs via the small GTPase ADPri...

Early dissemination, metastasis and therapy resistance are central hallmarks of aggressive cancer types and the leading cause of cancer-associated deaths. The EMT-inducing transcriptional repressor ZEB1 is a crucial stimulator of these processes, particularly by coupling the activation of cellular motility with stemness and survival properties. ZEB1 expression is associated with aggressive behaviour in many tumour types, but the potent effects cannot be solely explained by its proven function as a transcriptional repressor of epithelial genes. Here we describe a direct interaction of ZEB1 with the Hippo pathway effector YAP, but notably not with its paralogue TAZ. In consequence, ZEB1 switches its function to a transcriptional co-activator of a 'common ZEB1/YAP target gene set', thereby linking two pathways with similar cancer promoting effects. This gene set is a predictor of poor survival, therapy resistance and increased metastatic risk in breast cancer, indicating the clinical relevance of our findings.

Most mitochondrial proteins possess N-terminal presequences that are required for targeting and import into the organelle. Upon import, presequences are cleaved off by matrix processing peptidases and subsequently degraded by the peptidasome Cym1/PreP, which also degrades Amyloid-beta peptides (Aβ). Here we find that impaired turnover of presequence peptides results in feedback inhibition of presequence processing enzymes. Moreover, Aβ inhibits degradation of presequence peptides by PreP, resulting in accumulation of mitochondrial preproteins and processing intermediates. Dysfunctional preprotein maturation leads to rapid protein degradation and an imbalanced organellar proteome. Our findings reveal a general mechanism by which Aβ peptide can induce the multiple diverse mitochondrial dysfunctions accompanying Alzheimer's disease.

No abstract

The mitochondrial proteome comprises ~1000 (yeast)–1500 (human) different proteins, which are distributed into four different subcompartments. The sublocalization of these proteins within the organelle in most cases remains poorly defined. Here we describe an integrated approach combining stable isotope labeling, various protein enrichment and extraction strategies and quantitative mass spectrometry to produce a quantitative map of submitochondrial protein distribution in S. cerevisiae. This quantitative landscape enables a proteome-wide classification of 986 proteins into soluble, peripheral, and integral mitochondrial membrane proteins, and the assignment of 818 proteins into the four subcompartments: outer membrane, inner membrane, intermembrane space, or matrix. We also identified 206 proteins that were not previously annotated as localized to mitochondria. Furthermore, the protease Prd1, misannotated as intermembrane space protein, could be re-assigned and characterized as a presequence peptide degrading enzyme in the matrix.