Since I began my lab, I have been interested in the regulation of growth, the process through which cells and organisms accumulate mass and increase in size. This broad topic has been tackled from many different angles. I chose to focus on the mechanistic TOR (mTOR) pathway because of its emerging role as a central regulator of cell, organ, and organismal size and its deregulation in common human diseases such as cancer and diabetes. Our work on mTOR has yielded several surprising findings that have led us in directions I did not foresee even a few years ago. Having started with work on mTOR signaling, we are now pursuing a diverse set of projects, including studies on the response of adult tissue stem cells to nutritional states like obesity, the response of the liver to calorie restriction, and the role of energy and nutrient metabolic pathways in cancer. In addition, we continue to participate in the development of new technologies, as described below.
During the past few years, we discovered several components of the mTOR pathway and their molecular and cellular functions. We now know that the mTOR protein kinase is the catalytic subunit of two distinct multiprotein complexes, called mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2). These complexes anchor interconnected signaling pathways that sense diverse signals, such as nutrients and insulin, to regulate many fundamental processes, including cell growth, survival, and division (see model in Figure 1). In ongoing efforts, we have identified additional components of the pathway and are making progress toward understanding how mTORC1 and mTORC2 are activated and signal to downstream effectors.
One of the most interesting aspects of the mTORC1 pathway is that it is regulated by a large variety of upstream signals. Many of these stimuli signal through the tumor-suppressor complex (TSC1/TSC2), but this is not the case for amino acids. We identified the Rag family of GTPases as mTORC1-interacting proteins that are necessary and sufficient for signaling amino acid sufficiency to mTORC1. The Rag proteins affect mTORC1 signaling in an unexpected way. They do not directly stimulate the kinase activity of mTORC1 but rather mediate its translocation, in an amino acid–dependent fashion, to the surface of lysosomes. There, mTORC1 can interact with its well-established activator Rheb, which is regulated by signals such as insulin. This system is an elegant way for cells to prevent mTORC1 activation in the absence of amino acids, building blocks of cellular mass and important energy sources. We are now focusing on understanding how amino acids are sensed and have support for a model in which sensing originates within the lysosomal lumen through a mechanism that requires the vacuolar ATPase. We also recently identified guanine nucleotide exchange and GTPase activating protein (GAP) proteins for the Rag GTPases and are studying their regulation as well.
We are also using our insights into mTOR signaling to study normal physiology in vivo in mice. For example, by activating mTORC1 in vivo under conditions in which the pathway should be off, we are learning about the role of organ growth in the physiological response to calorie restriction. In addition, we have developed an interest in how normal tissue stem cells respond to nutritional states and have identified a pathway in the small intestine that increases the self-renewal of stem cells in response to calorie restriction.
Since my training as an M.D./Ph.D. student, I have wanted my research to have an impact on the understanding and treatment of human disease. With this goal, we are studying the roles of mTORC1 and mTORC2 in tumorigenesis, basing the work on our understanding of their biochemical functions within cells. A number of years ago, we discovered that mTORC2 is the elusive "hydrophobic motif kinase" of Akt and thus a positive regulator of the PI3K (phosphatidylinositol 3-kinase)/Akt pathway (Figure 1), a signaling system that is hyperactive in cancer cells missing the PTEN tumor suppressor. mTORC2-mediated phosphorylation of Akt is necessary for maximal but not basal Akt activity, suggesting that mTORC2 inhibition may selectively impair cells that depend on hyperactive PI3K/Akt signaling, such as those missing PTEN. Indeed, in vivo we find that mTORC2 is necessary for the formation of invasive prostate adenocarcinoma caused by PTEN loss but not for biogenesis of the normal prostate.
To generate tools to inhibit mTORC2 after mice have developed prostate cancer, we have a long-standing collaboration with Nathanael Gray (Dana-Farber Cancer Institute) to identify small molecules that inhibit mTOR or mTORC1 or mTORC2 selectively. Recently, we found that the genes encoding the trimeric GAP complex we identified for the Rag GTPases are tumor suppressors, so their loss may serve as a biomarker for predicting which tumors respond to existing mTOR inhibitors.
We are also undertaking systematic efforts to identify and understand the functions of the metabolic processes that are essential in tumor cells that carry common cancer-causing mutations. Our approach is to use loss-of-function screens, either with RNAi or by insertional mutagenesis in haploid human cells, of all the human metabolic genes to identify those necessary for cancer cells to survive in standard tissue culture conditions and in conditions that mimic oxygen-deprived regions of a tumor. In addition, we have recently set up metabolite profiling at the Whitehead Institute, which is greatly aiding our functional analysis of the genes we have identified.
During the past few years, our laboratory has been developing additional technologies to systematically screen for gene function in mammalian cells. Our work first led to a "reverse transfection" system that generates microarrays of mammalian cells expressing defined cDNAs. In this system, cells are cultured on a glass slide printed in known locations with nanoliter volumes of a hydrogel containing cDNAs in expression plasmids. The cells that land on the spots printed with the DNAs take up the plasmids to form clusters of 80–200 cells that express the proteins encoded by the plasmids. Because each spot is only 150–250 μm in diameter, we can print 5,000–10,000 such spots on one standard microscope slide. More recently, we developed analogous systems in which we create microarrays of purified lentiviruses expressing cDNAs or shRNAs.
Portions of this research were supported by the National Institutes of Health, W.M. Keck Foundation, Rita Allen Foundation, Pew Scholars Program, Department of Defense, Ellison Foundation, Whitehead Institute for Biomedical Research, Broad Institute, and David H. Koch Institute for Integrative Cancer Research at MIT.
As of April 02, 2013