Skip to Content

Garth Powis, D.Phil.

Chair 
Experimental Therapeutics
University of Texas MD Anderson Cancer Center
1515 Holcombe Blvd. Unit 0422, FC6.3044
Houston, TX 77030
(713) 745-3366
(713) 606-3695 pager
gpowis@mdanderson.org


My research activities fall into four main mechanistic and therapeutic areas:

  1. Redox Signaling – We are the first group to uncover a role for the thioredoxin reductase/thioredoxin (TR/Trx) system in cancer that shows many cancer cells overexpress Trx and its increased expression is correlated with aggressive tumor growth, decreased apoptosis and decreased patient survival. The TR/Trx system is involved in many aspects of cell function but its major effects in cancer are to inhibit the activity of the cell survival proteins apoptosis signal-regulating kinase 1 (ASK1) and, as we have shown, PTEN, and to regulate the activity of redox sensitive transcription factors. Most recently we conducted a forward genetic screen in Drosophila with mutant thioredoxin reductase uncovering known and novel pathways of redox signaling. One of the unexpected redox pathways we discovered is Wnt signaling that we subsequently show to be redox-dependent through TR/Trx regulation of TCF transcription factor activity in human colon cancer cells. This is the first indication of the redox regulation of Wnt signaling opening up new possibilities for therapeutic attack. We developed a small molecule inhibitor of Trx, PX-12, which has recently completed a Phase II clinical trial in pancreatic cancer. Our new studies show that PX-12 has antitumor activity in the APCmin-/+ mouse model of human familial adenomatous polyposis (FAP) that is driven by activated Wnt signaling. Ongoing studies investigate the role of TGFβ in inducing Trx levels in tumors as a response to stress.
  2. Hypoxia – Many solid tumors are hypoxic due to an abnormal tumor vasculature as well as areas of hypoxia around rapidly growing tumors. The main response to hypoxia is through the hypoxia inducible factors-1 and -2 (HIF-1/2) transcription factors that induce the expression of a number of genes that control cancer cell glycolysis, angiogenesis, cell survival and metastasis. Our current work focuses on the role of the hypoxia accessory protein (HAF) in inducing HIF-1 degradation while stabilizing and activating HIF-2. Thus, increased HAF, which is seen in many human tumors, is a switch from HIF-1 to HIF-2 and is important because HIF-2 is found in cancer progenitor (stem) cells that are found in hypoxic niches in the body. HIF-2 is known to induce progenitor cell genes. We have shown that HAF induces cell stemness in glioblastoma giving highly aggressive tumors. Another important finding is that hypoxia acting through HIF-1 is responsible for the activation of hedgehog signaling in pancreatic cancer cells leading to the characteristic inflammatory fibrosis (desmoplasia) of pancreatic and other cancers. Hypoxia increases sonic hedgehog ligand production by tumor cells that acts on stromal myofibroblasts to increase their hedgehog signaling by activating the GLi-1 transcription factor that causes the production of stromal collagen and, thus, desmoplasia. Increased intratumoral pressure caused by desmoplasia decreases blood flow, leading to increased hypoxia and further desmoplasia. We have shown that HIF-1 dependent increased sonic hedgehog levels are associated with decreased patient survival in pancreatic cancer. An inhibitor of HIF-1/2 translation that we developed, PX-478, is in Phase I clinical trial. Among its effects we have shown is the inhibition of hedgehog signaling, marked antitumor activity in animal models and potentiation of the effects of radiation.
  3. PI-3-Kinase Signaling – My research laboratory is a pioneer in the study of PI-3-kinase/AKT signaling. We have an irreversible PI-3-kinase inhibitor, PX-866, in Phase I clinical trial. PX-866 is a stabilized analog of the PI-3-K inhibitor wortmannin, which we first discovered. We have shown that PX-866 inhibits cell motility at subnanomolar concentrations and inhibits the growth of tumors with activated PI-3-kinase signaling. Significantly, we also have shown that mut-KRAS overrides the effects of this PI-3-kinase inhibition so that patients with mut-KRAS tumors are unlikely to be responsive to PI-3-kinase therapy alone. As an extension of this work, and working with clinical investigators of the BATTLE-1 program, we have shown that while mut-KRAS alone does not predict for patient survival different mut-KRAS amino acid substitutions predict patient survival differently, with mut-KRAS Cys predicting for decreased patient survival compared to other mut-KRAS amino acid substitutions. Furthermore, we have shown that different mut-KRAS have different structural configurations in the critical Switch II region allowing mutant forms to act differently with different downstream signal transducers – mut-KRAS Cys with RAL and mut-KRAS Asp with PI-3-kinase
  4. Pleckstrin homology (PH) domain inhibitors – The PH domain is a 3-dimensional protein fold found in many signaling proteins that binds to PI-3-phosphates, thus directing the host protein to the inner plasma membrane where they are activated by other recruited PH domain kinases. We were the first group to demonstrate the druggability of the PH domain by the PI-3-P analog 3-deoxyy PI-ether lipid (DPIEL). Building on this work, and using computational docking of large (3 million plus) virtual chemical libraries, we are developing PH domain inhibitors of a number of signaling proteins including AKT, PDPK1 and, most recently, a mut-KRAS nanocluster protein that specifically regulates mut-KRAS activity.

© 2014 The University of Texas MD Anderson Cancer Center