Kinase Fusion proteins, produced by translocations or other chromosomal rearrangements, represent an important class of oncogenes that are viable targets for the development of anti-cancer drugs. Our work in this area includes projects such as DNAJB1-PKACA as the driver of Fibrolamellar hepatocellular carcinoma and ALK fusion proteins in cancers. A. DNAJB1-PKACA Fibrolamellar hepatocellular carcinoma (FLHCC) is a liver cancer that predominantly affects adolescents and young adults. Effective therapeutic options are very limited because FLHCC does not respond well to chemotherapy. Given the low survival rate and lack of available treatment options, there is a pressing need for new diagnostic and therapeutic approaches to FLHCC. Inhibition of the tumor driver has the best chance of offering a curative treatment for FLHCC. The results of whole genome sequencing and transcriptome sequencing of FLHCC tumors show a single, consistent genetic alteration in the FLHCC tumor cells: a deletion of 400 kB between the first exon of the heat shock protein DNAJB1 and the first exon of the catalytic subunit of Protein Kinase A (PKA), PKACA, on one copy of chromosome 19. This deletion produces a chimeric gene that leads to a chimera protein. This chimeric DNAJB1-PKACA protein has now been recognized as the driver of FLHCC [1]. Inhibition of the DNAJB1-PKACA chimeric tumor driver offers tremendous potential to cure FLHCC. My group will study the structure of this chimera and how it is regulated with the long-term goal of developing precision medicines against this fatal pediatric cancer. This DNAJB1-PKA CA chimera, referred to as liverC (LC), forms a stable holoenzyme complex with R (R2:LC2) . We will focus on RIA and RIIB holoenzymes in this study since these are the two major PKA R isoforms found in FLHCC tumor cells.
This aim has three specific research objectives: A1) To characterize the structure and function of a R2:LC2 holoenzyme. LC is a very stable protein that can be easily expressed in E. coli and forms a stable complex with the R homodimer. We plan to solve the atomic resolution structures of the RIA2:LC2 and RIIB2:LC2 holoenzymes by using both x-ray crystallography and cryoEM techniques in parallel. I have successfully assembled the R2:LC2 complexes, used these complexes for crystallization trials and obtained promising crystallization conditions. Crystal optimization experiments are under way. We are varying pH, buffer, protein concentration, precipitant concentration, seeding, and additives to improve the crystal quality. Once we obtain well-diffracting crystals, we will solve and analyze the structure. I have also obtained very promising two dimensional single -particle cryo-EM images of the RIA2:LC2 complexes (molecular weight 180kDa). Experiments for complex particle stabilization under cryo conditions are under way. A2) Structure determination of a heterotetramer containing R2 dimer, wt C, and liver C. To determine if the LC fusion protein is a dominant mutation we formed a hetero-tetramer by adding a 1:1:1 mixture of the R2 dimer, wt C and LC (R2:C:LC). Surprisingly, formation of the hetero-tetramer appeared to be dominant based on our gel filtration results , which showed little of the wt holoenzyme or holoenzyme formed with two LC subunits. We are poised now to use the R2:C:LC for both x-ray crystallography and cryoEM experiments. In parallel, we will carry out biophysical and biochemical characterizations of the holoenzyme. Most FLHCC patients also have equimolar concentrations of wt C and LC-subunits suggesting that the heterodimer may be prevalent and likely dominant. A3) To perform virtual screening for inhibitors of LC and R2:LC2. I propose to perform structure-based virtual screening on vast compound libraries for inhibitors of LC and R2:LC2 holoenyzmes to identify putative inhibitors of either the catalytic pocket or an allosteric site. Likely binders will be tested for selective inhibitory activity via kinase assays. The identified small molecules should be able to selectively inhibit the chimera LC over the wt C or should selectively inhibit the activation of chimera R2:LC2 holoenyzmes over the R2:C2 holoenyzmes. B. ALK Non-small cell lung cancer (NSCLC) is a major cause of death worldwide, with most of the patients being diagnosed with disease in advanced stage. This disease also has a key driver in the form of a kinase fusion protein. The fusion between echinoderm microtubule-associated protein-like 4 (EML4) and anaplastic lymphoma kinase (ALK) has been identified in a subset (7%) of NSCLCs, equivalent to over 70,000 patients diagnosed annually worldwide. EML4-ALK occurs most often in non-smokers with lung cancer and is oncogenic both in vitro and in vivo . All EML4-ALK fusions contain a coiled-coil domain within EML4 that mediates constitutive dimerization and activation of EML4-ALK. NSCLCs harboring rearrangements involving the ALK gene are sensitive to treatment with the ALK inhibitor drug crizotinib, which is an ATP analogue. However, the enormous success of traditional kinase ATP analogue inhibitors in the treatment of cancer is being rapidly overshadowed by the emergence of drug resistance. Drug-resistant mutations increase the kinases' binding affinity with ATP, shifting them to the active conformation, and/or introduce new steric hindrance that interferes with the inhibitor structural motifs outside the highly conserved ATP binding boundary. Overall, the traditionally designed, oversized kinase inhibitors are destined to develop many different resistance profiles in the clinic. Continuously introducing new inhibitors to overcome new evolving mutations is becoming increasingly hard to sustain. Targeting resistance-driven kinase active conformation with a small compact molecule that is completely located inside the ATP binding boundary offers tremendous potential to tackle the ever evolving mutation resistances. Research objectives. We will achieve the goal of generating novel strategies to target ALK through three research objectives. B1) Solve the structure of EML4-ALK. EML4-ALK fusion protein is a homodimer with a molecular weight of 220kDs in total and is constitutively active. Our goal is to capture the active conformation of ALK using both x-ray crystallography and cryoEM approaches. B2) Solve the structure of EML4-ALK(G1202R), which confers a common resistance to crizotinib and other ALK inhibitors. B3) Explore EML4-ALK and EML4-ALK (G1202R)'s interactions with small compact molecules that are completely located inside the ATP binding boundary. The top three EML4-ALK variants (V1, V2 and V3) will be tested for expression in insect, bacterial, and mammalian expression systems followed by purification. These purified WT and mutant proteins will be used in the apo form or with drugs for crystallization screening and cryoEM trials. Other options include adding ATP or its analogues, as an alternative approach to increase our chance of success for obtaining high resolution structures. In parallel, we will carry out biophysical and biochemical characterizations of the EML4-ALK to elucidate its oligomerization state, ATP and substrate binding, and activity in pathological states and further test structure-guided hypotheses by using an appropriate functional assay. We hope to combine structural and functional studies to reveal the molecular mechanism of dysfunction of ALK kinase complexes and to help develop new strategies for structure based drug design.
