Abberant kinase-depenent signaling is associated with the etiology of several cancers. For this reason, pharmacological agents are being developed to modulate kinase-dependent signaling as potential new anticancer therapeutics. We are developing kinase-dependent signaling inhibitors that function by: (1) Blocking protein-protein associations mediated by recognition and binding of the polobox binding domain (PBD) of polo-like kinase 1 (Plk1) to phosphothreonine (pThr) and phosphoserine (pSer)-containing protein sequences and (2) Blocking the removal of phosphoryl groups by cellular protein-tyrosine phosphatases (PTPs). (1) Polo-like Kinase 1 (Plk1) Polo Box Domain Binding Inhibitors: Overexpression of the serine-threonine polo-like kinase 1 (Plk1) is tightly associated with oncogenesis in several human cancers. Interference with Plk1 function induces apoptosis in tumor cells but not in normal cells. Accordingly, Plk1 is a potentially attractive anticancer chemotherapeutic target. Plk1 possesses a unique phosphopeptide binding polo box domain (PBD) that is essential for its intracellular localization and mitotic functions. Unlike kinase domains, PBDs are found only in the four members of Plks. Therefore, they represent ideal targets for selectively inhibiting the function of Plks. By examining various PBD-binding phosphopeptides, our NCI collaborator, Dr. Kyung Lee, previously found that the 5-mer phosphopeptide PLHSpT specifically interacts with the Plk1 PBD with high affinity, whereas it fails to significantly interact with the PBDs of two closely related kinases, Plk2 and Plk3. Starting from this peptide, we employed an iterative sequential process of structural refinement to arrive at new agents, which bind with high affinity to the Plk1 PBD. Several of these agents can inhibit binding interactions with the Plk1 PBD at concentrations (low nanomolar) that are 10,000 times more potent than the parent PLHSpT peptide. These peptides retain high selectivity for the Plk1 PBD relative to the related Plk2 or Plk3 PBDs. In collaboration with Dr. Michael Yaffe (MIT) X-ray cocrystal structures of these peptides bound to Plk1 PBD indicate unanticipated modes of binding that take advantage of a cryptic binding channel that is not present in the nonliganded PBD. In further work, we have examined a range of amino acid analogs designed to take advantage of the newly discovered cryptic binding channel. Several of these amino acid analogs exhibit significantly increased inhibitory potencies relative to the parent PLHSpT peptide. The binding modes exhibited by these inhibitors define a new genre of PBD-binding interactions that could greatly impact the field of PBD-directed inhibitors. Further work has been directed at modification of the pThr residue, which forms a key component of ligand recognition. Although critical elements in the high affinity recognition of peptides and proteins by PBD are derived from pThr or pSer residues, the use of these residues in therapeutics is potentially limited by their lability in the presence of cellular phosphatases and by their poor cellular uptake due to their high anionic charge. To date, there has been little examination of pThr or pSer replacements within a PBD context. Accordingly, we have investigated the abilities of a variety of amino acid residues and derivatives to serve as pThr or pSer replacements. This work has shed new light on structure activity relationships for PBD recognition of phosphoamino acid mimetics. Of particular note, is our discovery of mono-anionic pThr analogues that bind to the Plk1 PBD with single digit inhibitory potencies, and which exhibit enhanced efficacies in cellular assays. We have extended these findings by applying bio-reversible phosphoryl prodrug protection that yielded uncharged species with further enhanced potencies in cellular assays. (2) PTP Inhibitors: While our work to date has focused on the protein-tyrosine phosphatase (PTP) family, given their importance in cancer, we are initiating work to investigate the structural biology of a family of dual specificity phosphatases (DUSPs), which hydrolyze both pTyr and pThr or pSer-containing substrates. . This work employs a panel of phosphatases assembled by our collaborators, Drs. Waugh and Ulrich. Among the enzymes in this panel are VH1 (associated with the Variola virus, which causes small pox);VHR (Vaccinia H1-related, which is upregulated in several cancers);DUSP12 (glucokinase-associated dual specificity phosphatase);DUSP14;DUSP22;DUSP27, CDC25C and PRL-3 (also known as PTP4A3). The majority of the DUSPS were chosen for their oncogenic connections. X-ray crystallographic determination of structures is being undertaken by Dr. Waugh, and for several of the DUSP family members, these will represent the first reported structures. We are investigating factors that influence differential substrate affinity for each phosphatase using a tethered oxime library approach. We have demonstrated the utility of this approach in various contexts. Starting from a parent peptide sequence, which shows good affinity across several members of the family of phosphatases, we sequentially introduced an aminooxy-containing residue at every position of an N-terminally biotinylated parent peptide. Evaluating libraries of oximes at each residue position represents a tethered fragment approach, which allows exploration of structural diversity significantly beyond what would be attainable using coded amino acids. In work performed in the laboratory of Dr. Ulrich, our libraries of oxime-containing peptides were screened for their substrate activities against the panel of phosphatases. The determination of relative substrate efficacies was achieved using microarray techniques involving printing of the oxime-containing peptides onto avidin-containing slides (FAST slides). Each slide contained approximately 1000 peptides. The printed phosphopeptide arrays were subjected to phosphatase hydrolysis by each member of the panel of phosphatases, and the relative levels of non-hydrolyzed peptides remaining were then visualized using ELISA techniques. It is anticipated that data from these studies should facilitate the design of peptide mimetic inhibitors directed against members of the phosphatase panel. Finally, to compliment this work we are developing proteins that merge properties of antibodies with biologically active small molecules. This work is being done in collaboration with Dr. Christoph Rader (Scripps Florida). Our approach employs monoclonal antibodies and antibody Fc fragments harboring a single C-terminal selenocysteine residue (Fc-Sec). The resulting antibody drug conjugates (ADCs) are directed against a variety of targets by changing the peptide or small molecule to which they are conjugated. In one aspect of our work, we have employed a variety of chemistries to attach biologically cleavable linkers that allow release of cargo once delivery to the target has been achieved. We have developed versatile hetero bifunctional linkers incorporating biologically cleavable bonds that are compatible with multiple types of Cu-free Huisgen 1,3-dipolar cycloaddition reagents. These linkers contain both targeting functionality and drug payloads. In one aspect of our work involving the potently cytotoxic peptide, monomethyl auristatin F (MMAF), we are examining bio-cleavable linkers that can be conjugated to the Fc-Sec protein by nucleophilic alkylation reactions. This work has involved developing new synthetic routes to key components of the MMAF peptide.
Showing the most recent 10 out of 48 publications
