The activity of protein kinases are naturally gated by a variety of physiochemical inputs, such as phosphorylation, metal ions, and small molecules. In order to design protein kinases that can be gated by user-defined inputs, we describe a sequence dissimilarity based approach for identifying sites in protein kinases that accommodate 25-residue loop insertion while retaining catalytic activity. We further demonstrate that the successful loop insertion mutants provide guidance for the dissection of protein kinases into two fragments that cannot spontaneously assemble and are thus inactive but can be converted into ligand-gated catalytically active split-protein kinases. We successfully demonstrate the feasibility of this approach with Lyn, Fak, Src, and PKA, which suggests potential generality.

Protein kinases phosphorylate client proteins, while protein phosphatases catalyze their dephosphorylation and thereby in concert exert reversible control over numerous signal transduction pathways. We have recently reported the design and validation of split-protein kinases that can be conditionally activated by an added small molecule chemical inducer of dimerization (CID), rapamycin. Herein, we provide the rational design and validation of three split-tyrosine phosphatases (PTPs) attached to FKBP and FRB, where catalytic activity can be modulated with rapamycin. We further demonstrate that the orthogonal CIDs, abscisic acid and gibberellic acid, can be used to impart control over the activity of split-tyrosine kinases (PTKs). Finally, we demonstrate that designed split-phosphatases and split-kinases can be activated by orthogonal CIDs in mammalian cells. In sum, we provide a methodology that allows for post-translational orthogonal small molecule control over the activity of user defined split-PTKs and split-PTPs. This methodology has the long-term potential for both interrogating and redesigning phosphorylation dependent signaling pathways.

We have previously reported that 8-phenyl-2'-deoxyguanosine derivatives (8PhGs) are able to extract metal cations from an aqueous phase into an organic phase. Herein we report on the ability of 8PhGs to transport metal cations across a bulk lipophilic liquid membrane. The experiments were performed using lithium, sodium, potassium, and strontium picrate salts with the parent lipophilic Gi, two isomeric 8PhG derivatives, cis-dicyclohexano-18-crown-6 (CD18C6) and [2•2•2] cryptand as reference compounds. The relative amounts of the picrate salts were measured by UV spectroscopy in both, the source phase and the receiving phase over a period of 24 h. The results show that the transport efficiency of the self-assembled ionophores formed by 8PhGs is either similar or superior to that of CD18C6, and in all but one case higher than the parent compound Gi. The varying efficiencies between the derivatives can be attributed to the stability (kinetic and thermodynamic) and the different molecularities of the supramolecules formed by these 8PhGs. The ease of the synthesis of 8PhGs, their anion independent assembly and the fact that the transport efficiency can be modulated as a function of the structure of the 8PhGs bode well for the use of such compounds in the development of novel antimicrobial agents and cation sensing devices.

Protein kinases are implicated in diverse signaling cascades and have been targeted with small molecules that typically bind the conserved ATP-binding active site. These inhibitors are often promiscuous and target multiple protein kinases, which has led to the development of alternate strategies to discover selective ligands. We have recently described a fragment-based selection approach, where a small-molecule warhead can be non-covalently tethered to a phage-displayed library of cyclic peptides. This approach led to the conversion of the promiscuous kinase inhibitor, staurosporine, into a selective bivalent inhibitor.