During the evolution of metazoans and the rise of systemic hormonal regulation, the insulin-controlled class 1 phosphatidylinositol 3OH-kinase (PI3K) pathway was merged with the primordial amino acid-driven mammalian target of rapamycin (mTOR) pathway to control the growth and development of the organism. Insulin regulates mTOR function through a recently described canonical signaling pathway, which is initiated by the activation of class 1 PI3K. However, how the amino acid input is integrated with that of the insulin signaling pathway is unclear. Here we used a number of molecular, biochemical, and pharmacological approaches to address this issue. Unexpectedly, we found that a major pathway by which amino acids control mTOR signaling is distinct from that of insulin and that, instead of signaling through components of the insulin͞class 1 PI3K pathway, amino acids mediate mTOR activation by signaling through class 3 PI3K, hVps34.insulin ͉ nutrients ͉ S6 kinase 1 ͉ endosomes ͉ PI3P

Impaired ribosome biogenesis is attributed to nucleolar disruption and diffusion of a subset of 60S ribosomal proteins, particularly ribosomal protein (rp)L11, into the nucleoplasm, where they inhibit MDM2, leading to p53 induction and cell-cycle arrest1–4. Previously, we demonstrated that deletion of the 40S rpS6 gene in mouse liver prevents hepatocytes from re-entering the cell cycle after partial hepatectomy5. Here, we show that this response leads to an increase in p53, which is recapitulated in culture by rpS6-siRNA treatment and rescued by the simultaneous depletion of p53. However, disruption of biogenesis of 40S ribosomes had no effect on nucleolar integrity, although p53 induction was mediated by rpL11, leading to the finding that the cell selectively upregulates the translation of mRNAs with a polypyrimidine tract at their 5´-transcriptional start site (5´-TOP mRNAs), including that encoding rpL11, on impairment of 40S ribosome biogenesis. Increased 5´-TOP mRNA translation takes place despite continued 60S ribosome biogenesis and a decrease in global translation. Thus, in proliferative human disorders involving hypomorphic mutations in 40S ribosomal proteins6,7, specific targeting of rpL11 upregulation would spare other stress pathways that mediate the potential benefits of p53 induction8.

The largest gene knock-down experiments performed to date have used multiple short interfering/short hairpin (si/sh)RNAs per gene. To overcome this burden for design of a genome-wide siRNA library, we used the Stuttgart Neural Net Simulator to train algorithms on a data set of 2,182 randomly selected siRNAs targeted to 34 mRNA species, assayed through a high-throughput fluorescent reporter gene system. The algorithm, (BIOPREDsi), reliably predicted activity of 249 siRNAs of an independent test set (Pearson coefficient r = 0.66) and siRNAs targeting endogenous genes at mRNA and protein levels. Neural networks trained on a complementary 21-nucleotide (nt) guide sequence were superior to those trained on a 19-nt sequence. BIOPREDsi was used in the design of a genome-wide siRNA collection with two potent siRNAs per gene. When this collection of 50,000 siRNAs was used to identify genes involved in the cellular response to hypoxia, two of the most potent hits were the key hypoxia transcription factors HIF1A and ARNT.