We recently reported that: 1) TRAF2 is targeted for proteasome-dependent degradation by Siah2 in stress conditions that precede stress-induced apoptosis; 2) TNFa treatment causes rapid TRAF2 ubiquitination in vivo, which is dependent on TRAF2’s own intact RING and zinc finger domains, as well as on Ubc13 activity. This TRAF2 ubiquitination induces its translocation to the non-ionic detergent-insoluble cellular fraction (including membrane rafts and cytoskeletal complexes), resulting in selective activation of JNK but not of NF-kB and p38. On the basis of our published observations, we hypothesized that TRAF2-dependent activation of diverse signaling events is contingent on selective and distinct post-translational modifications, which are deregulated in cancer cells.

TRAF2 phosphorylation: Phosphorylation is one of the main post-translational modifications that regulate protein-protein interaction and kinase cascades. In vivo ³²P-orthophosphate labeling analysis revealed that TRAF2 is highly phosphorylated in transformed (293T and HeLa) cells and such phosphorylation is induced by TNFa stimulation in non-transformed (MEF) cells. Two-dimensional phosphoamino acid separation demonstrated that both basal and inducible TRAF2 phosphorylation take place on serine residues. Phospho-peptide mapping experiments indicate that at least five serine residues are phosphorylated on TRAF2 upon TNFa stimulation. We have now mapped four serine phosphorylation sites and identified the candidate kinases that are involved in TRAF2 phosphorylation. Interestingly, these kinases turn out to be well known oncoproteins that have pivotal roles in cancer transformation and tumor progression. Although, numerous studies have established the critical role of TRAF2 in activation of RIP/IKK/NF-kB and MAP3K/JNK signaling pathways, there are no studies that have been conducted to characterize the regulation of TRAF activity by post-translational modifications. None of the putative phosphorylation sites of any of the TRAF family proteins have been identified yet. Therefore, our study, characterization of the role of TRAF2 phosphorylation in TNFa- and stress-induced JNK and IKK activation, will shed a new light on understanding the molecular mechanisms for TNFa-induced activation of diverse signaling pathways.

TRAF2 ubiquitination: Both JNK and IKK are activated by TNFa within 3 min. and inactivated 30 min. after stimulation. Therefore, TNFR-signaling complex is either disrupted or degraded 30 min. after TNFa treatment. A critical yet most obscure step in TNFa signaling is the TRAF2-mediated activation and bifurcation of JNK and IKK signaling pathways. We have reported that the ubiquitination dependent translocation of TRAF2 to membrane rafts results in JNK but not in IKK activation and that TRAF2-mediated IKK activation is accomplished before TRAF2 ubiquitination. Recently, TRAF6 was reported to act as an E3 ligase and forms non-canonical K63-linked polyubiquitination chains, which appear necessary for TRAF6-mediated NF-kB activation. However, neither the putative substrate nor the exact mechanisms by which K63-linked polyubiquitin-chain are capable of activating both IKK and MAP3K have been identified. Direct evidence for the role of TRAFs ubiquitination in TRAF mediated signaling will come from identifying the ubiquitination sites of TRAF proteins. However, until now, no ubiquitination sites for any TRAF family members have been identified. In collaboration with Dr. Junmin Peng of Emory University, we identified the TRAF2 ubiquitination site by mass spectrometry analysis of ubiquitinated form of TRAF2. Mutation of this lysine to arginine (TRAF2-KR) abolishes TRAF2 ubiquitination completely in vivo. We will utilize this mutant TRAF2-KR to analyze the role of TRAF2 ubiquitination in TRAF2-mediated activation of JNK and IKK pathways in TRAF2 null MEF cells. The merits of the TRAF2-KR mutant over the other RING finger domain deleted or mutated TRAF2 are that it is a single lysine mutant and contains intact RING, zinc and TRAF domains. These domains are critical for TRAF2 binding with other interacting proteins. Thus, these studies will directly demonstrate the role of TRAF2 ubiquitination in TNFa signaling.

2. Explore the molecular mechanism for oncoprotein- and stress-induced NF-kB activation and cancer cell resistance to stress-induced apoptosis. The evasion of programmed cell death (apoptosis) is one of the common hallmarks of all types of cancer cells. Our increasing understanding of the molecular evolution of cancer has revealed that oncogene-driven deregulated proliferation is, in fact, hard-wired to signals p53-dependent cell cycle arrest, senescence and/or apoptosis. Therefore, inactivation of p53 function or other effectors upstream or downstream of p53 signaling pathways is an essential step in the transformation of precancerous cells to malignant cells. Recent studies have demonstrated that just as inactivation of the p53 pathway is universal in neoplasia, the activation of the NF-kB pathway is essential for cancer cell progression and metastasis.

Almost all cellular stresses and oncogenic stimulations that activate p53 also activate NF-kB. Activation of NF-kB promotes cellular proliferation and inhibits apoptosis, therefore counteracting p53 and favoring cancer development. Although the mechanism underlying the loss of p53 function in cancer cells is becoming clear through intensive studies over last two decades, the mechanism underlying the constitutive activation of NF-kB in human tumors are still elusive. Genetic alterations such as rearrangements, amplifications and missense mutations are very rare in kinases that are involved in activation of NF-kB. Therefore, it is believed that the culprits responsible for this constitutive NF-kB activation in cancer cells are upstream positive or negative regulatory proteins, which usually serve to increase or limit the NF-kB response.

By performing a series of conventional phospho-mapping experiments in combination with the application of different kinase inhibitors, we identified candidate kinases involved in TRAF2 phosphorylation. These kinases are normally activated by growth factors and cellular stresses and constitutively activated in many types of cancer cells. Therefore, we hypothesize that TRAF2 phosphorylation by these kinases is a critical event that mediates stress- and oncoprotein-induced NF-kB activation in cancer cells. Thus, the systematic analysis of the role of these kinases in TRAF2 phosphorylation and TRAF2-mediated IKK activation will uncover a new regulatory mechanism for oncoprotein- and stress-induced NF-kB activation, providing a rationale for developing anti-cancer drugs.

3. Identify new JNK substrates using ATP analogue and analogue sensitive JNK mutant for in vitro kinase reaction. Persistent JNK activation by TNFa stimulation in IKKb or RelA deficient MEF cells was reported to be one of the major inducers of apoptosis independent of FADD-mediated caspase pathway. Recent studies in JNK-deficient mice have indicated that JNK1 and JNK2 exert opposite effects in tumor promotion. While JNK1^-/- mice were shown to be more susceptible to TPA-induced skin tumor development, JNK2^-/- mice exhibited a significant reduction in skin tumor development induced by TPA. However, the known JNK substrates are far from the level of explaining all JNK functions. The classical two-hybrid screening and the affinity binding approach were proven to be not successful in identifying JNK substrates, as the interaction between JNK and its substrates are transient, and similar to a hit and run. In collaboration with Dr. Keven Shorkat of UCLA, we developed a novel method to identify new JNK substrates. The basic idea of this method is as follows: 1) generate different type of ATP analogues; 2) generate ATP analogue sensitive JNK (JNK-as) by mutation of hydrophobic, big amino acid in APT binding pocket of JNK (e.g. Met¹⁰⁸ in b5 sheet and Leu¹⁶⁸ in b8 sheet) to small hydrophobic Ala or Gly; 3) screen ATP analogue by in vitro kinase reaction to identify the ATP analogue that has the highest affinity to JNK-as, but not to wt-JNK; 4) use this ATP analogue, JNK-as and whole cell extracts to perform in vitro kinase reaction; 5) separate phosphorylated substrates by 2D-dimentional electrophoresis; and then 6) sequence positive spots (JNK substrates) by mass-spectrometry analysis. Using this novel method, we identified hnRNP K as a new JNK substrate. By fractionation of cytoplasmic, nuclear, ER or mitochondria protein extracts, and using the combination of this novel method, we attempt to identify more JNK substrates in the near future. This will advance our understanding of JNK mediated signaling pathways and their implication in inflammation and stress induced organ damage and cancer development.