Focus Issue: TOR Signaling, a Tale of Two Complexes

The kinase target of rapamycin (TOR) participates in two interconnected signaling networks through its association with two distinct protein complexes, TOR complex 1 (TORC1) and TOR complex 2 (TORC2). Although TORC1 and TORC2 have distinct upstream regulatory inputs as well as distinct downstream effectors, their combined activity enables the coordination of cellular growth with the cell cycle, growth factors, and nutrients. Through its regulation of cellular metabolism and the cytoskeleton, TOR signaling contributes to pathological processes associated with aging, cancer, and neurodegenerative disease, which makes understanding the TOR network clinically important. Indeed, various drugs targeting TOR, including drugs that are specific for mTORC1, such as rapamycin and its derivatives, or drugs that inhibit TOR in either complex, such as the TORKinibs, are either in clinical use or under investigation for clinical application.

There are several challenges to unraveling the TORC1 and TORC2 pathways and defining the connections between TOR signaling and other cellular regulatory systems, such as endoplasmic reticulum (ER) stress and autophagy. These include limited pharmacological tools for specifically manipulating the TORC2 pathway, the existence of feedback loops within the TORC1 pathway, and the positive input from TORC2 to an upstream regulator of the TORC1 pathway. One approach involves the use of model organisms such as yeast. Yeast provide a powerful model for dissecting the TOR pathway and for understanding how it is integrated with other cellular regulatory systems. In yeast, there are two TOR proteins, Tor1 and Tor2, which function in the TORC1 and TORC2 complexes, respectively. Cardon et al. (Research Article in the Archives) describe a pathway in budding yeast that involves the PASK family of protein kinases and the metabolic enzyme Ugp1 through which yeast can compensate for loss of TORC2 activity. In a Perspective in the Archives, Shiozaki highlights how fission yeast cell growth is coordinated with cell division through the integration of the TOR and stress-activated MAPK (mitogen-activated protein kinase) pathway.

More details of the TORC1 than of the TORC2 pathway have been identified. TORC1 is activated by nutrients (in particular, amino acids) and growth factors, such as insulin. Amino acids stimulate TORC1 through a process involving the translocation of TORC1 to lysosomal membranes by the Ragulator-Rag complex and phospholipase D1 (see the Perspective in this issue by Wiczer and Thomas). Insulin stimulates TORC1 through activation of phosphoinositide 3-kinase (PI3K) by the adaptor insulin-receptor substrate (IRS). PI3K produces phosphatidylinositol 3,4,5-trisphosphate (PIP₃), which activates phosphoinositide-dependent kinase and thus activates the kinase Akt. Akt inhibits the dimer composed of tuberous sclerosis complex 1/2 (TSC1/TSC2), alleviating its repression of the guanosine triphosphatase Rheb and thereby freeing Rheb to stimulate TORC1 activity. Downstream targets of TORC1 include p70 ribosomal protein S6 kinase (p70S6K), the translation initiation regulator 4E binding protein (4E-BP), and the proline-rich Akt substrate PRAS40 (see the Insulin Signaling Pathway). At least two feedback loops exist in this system: Phosphorylation of PRAS40 by TORC1 alleviates its inhibition, thereby serving as a positive feedback mechanism, and phosphorylation of IRS by p70S6K destabilizes IRS, which inhibits PI3K-mediated activation of TORC1, serving as a negative feedback mechanism. The upstream regulators of TORC2 have been the subject of debate, and several models for its regulation have been proposed. Complicating the study of the two pathways is a connection from TORC2 to Akt; TORC2 phosphorylation enhances Akt activity and thereby promotes TORC1 activation.

In this issue, Dalle Pezze et al. (see the Perspective by Fingar and Inoki) applied a combined computational-experimental approach to test various proposed network structures for the TORC2 pathway in mammalian cells. Their results led to the proposal of a distinct PI3K-dependent pathway from the insulin receptor to mammalian TORC2 (mTORC2) that is independent of various components of the pathways mediating the activation of mTORC1. Acetylation, and the nutrient-responsive reversal of this modification by enzymes of the HDAC family, have been implicated in regulation of mTORC2 activity (see the 2012 Editors’ Choice by Foley). Acetylation is not limited to the TORC2 pathway, but has also been described by Kuo et al. (Research Article in the Archives) for TSC2; this posttranslational modification stabilizes TSC2 and consequently inhibits TORC1 activity.

Despite substantial insight into the regulatory mechanisms controlling protein translation and TORC1 activity, additional mechanisms and kinases that control protein synthesis continue to be identified. As discussed in a Podcast (see the Research Article by Hu et al. in the Archives), Glass’s lab found that the kinase MNK2 has a negative role in the regulation of the protein synthesis pathway in skeletal muscle under conditions that promote atrophy. The inhibition of protein synthesis mechanisms involved a kinase-independent interaction with mTOR and a kinase-dependent regulation of another kinase, serine-arginine–rich protein kinase (SRPK), which functioned in a pathway that elicited reduced phosphorylation of eukaryotic translation initiation factor 4G (eIF4G). Mounir et al. (Research Article in the Archives) identified another kinase, the endoplasmic reticulum (ER)–resident protein kinase PERK, which inhibited the protein synthesis machinery. PERK activity and phosphorylation of eIF2α were inhibited by Akt or by ER stress or oxidative stress. Inactivation of the PERK-eIF2α pathway rendered tumor cells more susceptible to death after inhibition of PI3K-Akt signaling. Thus, PI3K-Akt signaling can promote protein synthesis not only through activation of TORC1 but also through the inhibition of the PERK-eIF2αP pathway. Targeting the PERK-eIF2α pathway may improve the efficacy of cancer therapies that target PI3K-Akt signaling.

Since the identification of rapamycin in the late 1970s as an immunosuppressive agent, the importance of TOR signaling in cancer and in changes associated with aging and with side effects associated with other drugs has sparked hope that this pathway could be targeted for many different human pathological conditions. Several Editors’ Choice summaries and Research Articles highlight developments in understanding the connection between TOR signaling and aging: Foley (Editors’ Choice 2009) describes how rapamycin increased mouse life span, and Ray discusses a relationship between sestrins and TOR in flies, showing that loss of sestrins results in abnormalities similar to those seen in humans with a sedentary life-style. Chen et al. 2009 (Research Article in the Archives) reported that rapamycin reversed aging-related declines in hematopoetic stem cells and enhanced the immune response of old mice. Santini et al. (see the Perspective by Klann in the Archives) found that inhibition of mTORC1 with rapamycin reduced the dyskinesia associated with a treatment for Parkinson’s disease, a neurodegenerative disease associated with aging.

Autophagy is a process by which intracellular constituents are enclosed in multivesicular bodies for degradation in response to ER stress, reduced nutrients, or inhibition of the PI3K-Akt pathway. Autophagy either can help cells survive or can contribute to an autophagic death process, and the interplay between TOR signaling and autophagy has been implicated in various pathophysiological conditions. Sun et al. (Research Article in the Archives) reported a connection between TOR signaling and autophagic death triggered by the H5N1 avian influenza virus. TORC1 can inhibit autophagy by phosphorylating components of the autophagic machinery (see the Perspective by Chan in the Archives). TORC2 activity toward Akt is inhibited by ER stress (see Chen et al. 2011 Research Article in the Archives). Disruption of the pathway through which ER stress inhibits this function of TORC2 may contribute to cancer cell proliferation. Loss of the negative regulator TSC1 causes increased activation of mTORC1 signaling, and, in a Research Article in this issue, Menon et al. show that hyperactivated mTORC1 signaling contributed to spontaneously occurring liver cancer in mice. Defects in autophagy and unresolved ER stress were some of the cellular changes associated with chronic mTORC1 signaling that may have contributed to tumorigenesis. Treating mice with rapamycin prior to the appearance of tumors blocked the development of hepatocellular carcinoma, as well as the pathological changes that preceded tumor formation. Because mTORC1 signaling increases during obesity, a major risk factor for liver cancer, mTORC1 may link environmental factors, such as diet, to the risk of developing certain types of cancers. Activation of autophagy may also enable cancer cells to survive therapies targeting the PI3K-Akt pathway. Fan et al. (Research Article in the Archives) explored the effects of various combinations of kinase and autophagy inhibitors on glioma cell survival and found that combined inhibition of mTORC1, PI3K, and autophagy, or that of mTORC1, mTORC2, and autophagy, triggered apoptosis. Because the drugs used in this study are either now in use in patients or in clinical trials, this suggests that the combination approach described could be readily translated into human therapy.

Identifying combination therapies for cancer treatment should be facilitated by understanding the changes in the signaling networks that occur in response to drug treatment. Moritz et al. (Research Article in the Archives) describe a phosphoproteomic approach to identify key proteins that exhibit changes in phosphorylation in response to inhibition of the PI3K-Akt pathway, the Ras to MAPK (mitogen-activated protein kinase) to RSK (ribosomal S6 kinase) pathway, and the mTORC1 to p70S6K pathway. In a Research Resource in this issue, Dephoure and Gygi describe a technique that combines two different labeling methods to enable the simultaneous analysis of samples from multiple conditions by mass spectrometry. Using this method, Dephoure and Gygi monitored changes in protein abundance in yeast in response to rapamycin to produce a data set that should further the discovery of the TOR network and how inhibition of TORC1 alters cellular behavior.

The studies highlighted here describe diverse approaches for discovering the intricacies of the TOR signaling network and emphasize the physiological importance of both TORC1 and TORC2. Understanding the details of this complex network and its intersection with other cellular systems should enable more specific manipulation of the pathway for clinical benefit.