Inhibitor of apoptosis proteins (IAPs) are important regulators of apoptosis and pro-survival signaling pathways whose deregulation is often associated with tumor genesis and tumor growth. IAPs have been proposed as targets for anticancer therapy, and a number of peptidomimetic IAP antagonists have entered clinical trials. Using our fragment-based screening approach, we identified nonpeptidic fragments binding with millimolar affinities to both cellular inhibitor of apoptosis protein 1 (cIAP1) and X-linked inhibitor of apoptosis protein (XIAP). Structure-based hit optimization together with an analysis of protein-ligand electrostatic potential complementarity allowed us to significantly increase binding affinity of the starting hits. Subsequent optimization gave a potent nonalanine IAP antagonist structurally distinct from all IAP antagonists previously reported. The lead compound had activity in cell-based assays and in a mouse xenograft efficacy model and represents a highly promising start point for further optimization.

XIAP and cIAP1 are members of the inhibitor of apoptosis protein (IAP) family and are key regulators of anti-apoptotic and pro-survival signaling pathways. Overexpression of IAPs occurs in various cancers and has been associated with tumor progression and resistance to treatment. Structure-based drug design (SBDD) guided by structural information from X-ray crystallography, computational studies, and NMR solution conformational analysis was successfully applied to a fragment-derived lead resulting in AT-IAP, a potent, orally bioavailable, dual antagonist of XIAP and cIAP1 and a structurally novel chemical probe for IAP biology.

A cyclam-based macrocyclic sensor has been prepared using synthetically simple "click" chemistry to link a fluorophore to the macrocyclic receptor. This sensor shows high selectivity for Zn(II) over a range of other metals, providing a significant enhancement of fluorescence intensity over a wide pH range. As such, this is the first cyclam-based sensor demonstrated to be selective for Zn(II) and is the first example of a triazole being used as a coordinating ligand on an azamacrocycle. The sensor can access biologically available zinc in mammalian cells, sensing the Zn(II) flux that exists during apoptotic cell death.

In an effort to improve upon the recently reported cyclam based zinc sensor 1, the "click"-generated 1,8-disubstituted analogue 2 has been prepared. The ligand shows a 2-fold increase in its fluorescence emission compared to 1 exclusively in the presence of Zn(II) that is typical of switch-on PET fluorescent sensors. Single crystal X-ray diffraction of complexes of model ligand 10 reveals that the configuration adopted by the macrocyclic framework is extremely sensitive to the metal ion to which it coordinates. For Zn(II), Mg(II), and Li(I) the metal ions adopt an octahedral geometry with a trans III configuration of the cyclam ring. In contrast for Ni(II) the ligand adopts the rare cis V configuration, while for Cu(II) a clear preference for five-coordinate geometry is displayed with a trans I configuration of the macrocyclic ring being observed in two essentially isostructural compounds prepared via different routes. The ligand displays an increased selectivity for Zn(II) compared to 1 in the majority of cases with excellent selectivity upheld over Na(I), Mg(II), Ca(II), Mn(II), Ni(II), Co(II), and Fe(III). In contrast for Cu(II) and Hg(II) little improvement was observed for 2 compared to 1 and for Cd(II) the selectivity of the new ligand was inferior. In the light of these findings and the slower response times for ligand 2, our original "click"-generated cyclam sensor system 1 was employed in a proof of concept study to prepare a heterogeneous sol-gel based material which retains its PET response to Zn(II). The versatile nature of the sol-gel process importantly allows the simple preparation of a variety of nanostructured materials displaying high surface area-volume ratio using fabrication methods such as soft lithography, electrospinning, and nanopipetting.

Chemical sensing is a mature field, and many effective sensors for small anions and cations have been devised. Metal complexes have been used widely for this purpose, but there are fewer reports of their use in the detection of organic and biological analytes. To date metal complexes have been used in sensing via the direct displacement of a pre-existing ligand by an analyte, or by an adventitious complementarity between the complex and analyte. These strategies do not permit a general approach to the sensing of biological molecules with metal complexes because of the demands to engineer molecular recognition into the complex architecture. We describe a fundamentally new approach to this field-the "allosteric scorpionate" metal complex. The binding partner of a biological analyte is attached to a scorpionate ligand on a metal complex, remote from the metal centre. Binding of the analyte causes a change in the primary coordination sphere at the metal, thereby revealing the presence of the biological molecule. We show that azamacrocyclic complexes with a triazole scorpion ligand may be easily assembled with the [3+2] Huisgens 'click' cycloaddition. We demonstrate the synthesis of a biotin-functionalised cyclam derivative using this methodology. This, and our previously communicated zinc sensor, are to the best of our knowledge the first examples of a triazole being employed as a scorpion ligand on an azamacrocycle. Coordination by the triazole to the metal is perturbed by the binding of avidin to the pendant ligand. This event can be sensitively detected with EPR spectroscopy, and the details of the coordination change probed with ENDOR spectroscopy, confirming the loss of the axial triazole nitrogen donor upon binding to avidin. This represents the first metal complex where remote, 'allosteric' coordination of an analyte has been shown to cause a change in the primary coordination sphere of the metal. Since the synthesis is modular and straightforward, other biological ligands may easily be introduced, and the associated binding events may be probed.

The members of the NSD subfamily of lysine methyl transferases are compelling oncology targets due to the recent characterization of gain-of-function mutations and translocations in several hematological cancers. To date, these proteins have proven intractable to small molecule inhibition. Here, we present initial efforts to identify inhibitors of MMSET (aka NSD2 or WHSC1) using solution phase and crystal structural methods. On the basis of 2D NMR experiments comparing NSD1 and MMSET structural mobility, we designed an MMSET construct with five point mutations in the N-terminal helix of its SET domain for crystallization experiments and elucidated the structure of the mutant MMSET SET domain at 2.1 Å resolution. Both NSD1 and MMSET crystal systems proved resistant to soaking or cocrystallography with inhibitors. However, use of the close homologue SETD2 as a structural surrogate supported the design and characterization of N-alkyl sinefungin derivatives, which showed low micromolar inhibition against both SETD2 and MMSET.

The DDR1 and DDR2 receptor tyrosine kinases are activated by extracellular collagen and have been implicated in a number of human diseases including cancer. We performed a fragment-based screen against DDR1 and identified fragments that bound either at the hinge or in the back pocket associated with the DFG-out conformation of the kinase. Modeling based on crystal structures of potent kinase inhibitors facilitated the "back-to-front" design of potent DDR1/2 inhibitors that incorporated one of the DFG-out fragments. Further optimization led to low nanomolar, orally bioavailable inhibitors that were selective for DDR1 and DDR2. The inhibitors were shown to potently inhibit DDR2 activity in cells but in contrast to unselective inhibitors such as dasatinib, they did not inhibit proliferation of mutant DDR2 lung SCC cell lines. KEYWORDS: discoidin domain receptor, fragment-based drug design, back to front kinase design D iscoidin domain receptors, DDR1 and DDR2, are transmembrane receptor tyrosine kinases that are activated by collagen binding to their extracellular domain. 1,2 DDR1 and DDR2 have been associated with extracellular remodeling, cell adhesion, proliferation and migration, and they have been linked to a number of human diseases, including fibrotic disorders, atherosclerosis and cancer. 3−5 There has recently been evidence suggesting that DDR2 inhibitors would be useful for the treatment of lung squamous cell cancer 6,7 although the role of DDR2 may be more complex than first realized. 8−10 Hammerman et al. 6,7 have shown that DDR2 is mutated in approximately 4% of lung squamous cell cancer and have presented data to suggest that these are gain of function mutations. Hammerman et al. 6 have also shown that cell lines harboring these mutations are selectively sensitized through knockdown of DDR2 by RNA interference or by treatment of the multitargeted kinase inhibitor dasatinib. 11 Selective DDR2 inhibitors would therefore be useful to probe the role of DDR2 and may have utility for the treatment of lung cancer. To date, most DDR1/2 inhibitors have been derived from cross-screening of existing kinase inhibitors. 12 Initial compounds often lacked selectivity over homologous enzymes such as Bcr-Abl, but a number of recent papers have described more selective inhibitors which all bind to the DFG-out form of the enzyme. 13−15 Gray and co-workers have identified DDR1/2 inhibitors by screening a library of compounds that had previously been designed by mixing and matching motifs from known Type II kinase inhibitors. 14 The resulting compounds were reported to be selective with IC 50 values versus DDR1 and DDR2 of approximately 100 nM, and experimental binding modes in DDR1 were determined. 14,16 Gao et al. have published a series of potent Type II DDR1 inhibitors that are selective over other kinases, including DDR2. 13 In this Letter we describe a fragment based approach to the discovery of potent and selective DDR1/2 inhibitors. We first obtained a soakable crystal form of DDR1 suitable for high-throughput cry...

We describe herein the remarkable, protoninduced transformations of supramolecular "chameleons" based on naphthalenediimides (NDIs). We demonstrate rapid, reversible, and controllable morphological switching between receptors for different fullerenes (C 60 and C 70 ), thus allowing the selective binding of either guest in a mixture of both guests. This work is an extension of the dynamic combinatorial concept [1,2] into a new dimension: using hydrogen bonding as the exchange reaction, the response of the NDI building blocks to the presence of fullerene guests depends on the concentration of protons as a third component. The switching between the two receptors, a nanotube and a hexameric capsule (Figure 1), is under thermodynamic control (i.e., the most stable host-guest complex is dominant) and is triggered by the guest (template) present in solution.In aprotic solvents of medium polarity such as chloroform and dichloromethane, self-recognition through hydrogen bonding causes the amino acid derived NDIs (Scheme 1) to adopt different aggregate forms, depending on the presence or absence of guests. [3][4][5] The NDI nanotubes are held together by classical intermolecular COOH-HOOC hydrogen bonds supplemented by weak CH···O = C bonding (Figure 1 a, c).[3] The hexameric capsule is formed in the presence of C 70 at the expense of the nanotube and is held together by hydrogen bonds between the COOH groups (equator, Figure 1 b) and a rare, threefold hydrogen-bond pattern between the COOH group, an imide C=O group, and an acidic hydrogen atom of NDI at the pole (Figure 1 d). [5] Both of these supramolecular assemblies can therefore be destroyed by offering the NDI units alternative hydrogenbonding interactions. We originally showed that this goal could be achieved by use of a hydrogen-bond-disrupting solvent such as MeOH; this approach has the disadvantage of leading to the permanent destruction of the supramolecules. We now show that morphological switching between nanotube, hexameric receptor, and monomers is readily achieved by simple protonation-deprotonation reactions that result in the formation of dynamic, size-selective fullerene receptors, the structure and recognition properties of which depend on the position of the acid/base equilibrium (Figure 2). This work has also uncovered unexpected differences in the sensitivity to base-induced dissociation of the nanotubes derived from different amino acids. Scheme 1. Naphthalenediimides derived from amino acids used in the present work. Boc = tert-butyloxycarbonyl, Bzl = benzyl, Trt = trityl.