Lithium Citrate australia The compounds f and a e were found
The compounds 8f and 9a–e were found to be potent inhibitors of both isoforms of GSK-3 as characterized by IC50 values in the low nanomolar range. In addition, all of them showed good selectivities against other kinases. Substituents in the ortho- and para-positions of structure 8 and 9 lead to potent inhibition of both isoforms as indicated for Lithium Citrate australia 8f and 9e with an IC50 value of 5 nM for GSK-3α respectively 14 nM and 32 nM for GSK-3β. We reasoned that only the interplay respectively combination of different substituents is adequate to gain selectivity against one GSK-3 isoform. Compounds 8b and 8g enhance the selectivity for GSK-3α up to 27-fold. An inverse effect was observed for pyridine 10e where a substituent in meta- and para-position leads to slightly increased selectivity for inhibition of GSK-3β. All compounds lacking the heterocycle (Scheme 3) or oxadiazole ring (Scheme 4) completely lost most of their ability to inhibit GSK-3 and resulted in a residual GSK-3β activity of more than 50% at 10 μM. We attempted the combination of different lead structures by the addition of an amide function at the pyridine and benzothiazole (Scheme 4, Scheme 5). A similar moiety serves as an ambident moiety on Sorafenib. It was reported to establish hydrogen bonds with the same amide on the target hinge region either via the pyridyl nitrogen or through the carbonyl of the acetyl amide , . Thus the amide was expected to increase the interaction with the backbone of GSK-3 as observed in the tight binding of Compound 62 and AR-A014418 (Fig. 1). We determined promising activities and selectivities for the pyridines 26a, 26c and 26d. Especially compound 26d with an IC50 value of 2 nM for GSK-3α and 17 nM for GSK-3β showed more than 5000-fold selectivity against Cdk5/p35, CK1ɛ, AurKA and PKCα (Table 1). However, the additional amide on the benzothiazole 35 reduced activity for both isoforms indicating too close proximity to the backbone. Moreover, the acetamide 26d is characterized by a clogP of 2.90 and a tPSA of 103.0 Å2. The letter value exceeds the limits for likely blood–brain barrier permeation. In addition, the structural rigidity of the dibenzofuran 41 resulted in loss of inhibitory activity against GSK-3. A docking study of compound 9e and 26d into the PDB structure 3F88 of GSK-3β suggested a binding mode that uses the ATP-binding pocket in its entirety. The oxadiazole interacts in both cases with the polar pocket consisting of Lys85 and Asp200. For compound 9e the amide functions leads to a binding mode along the hinge region of the ATP-binding pocket. The biphenyl system interacts with Ile62, Gly63, Phe67 and Val70, which form part of the flexible glycine-rich loop. The methoxy group of compound 9e is located in close proximity to the salt bridge formed by Glu137 and Arg141, which may be responsible for the activity of this compound. Nevertheless, this structure does not provide insight into the selective inhibition of the GSK-3β isoforms as all amino acid residues shown in Fig. 3 are identical in both isoforms. Hence, only a complex of GSK-3β co-crystallized with one of these inhibitors, solved by X-ray crystallography, may provide the necessary insights. After evaluation of the in vitro activity and selectivity of these compounds the biphenyl derivatives 9e and 26d were.further tested for their in vivo activity on wt zebrafish embryos. We exposed the embryos to these compounds at early stages of development (Fig. 4). The embryos were collected and maintained in E2 medium at ∼28 °C. The compounds were added 5 hpf (hour post fertilization), and the phenotypes were compared at 44-48 hpf. Compounds 9e and 26d cause a stunted and crooked tail at 30 μM. This correlates with the observation that Wnt signaling, and thus GSK-3β activity, plays a crucial role in the development of metazoan. This was demonstrated for the known GSK-3 inhibitors LiCl and the ruthenium complex (R)-7, which both perturb zebrafish development , . However, it must be noted that compound 9e and 26d were not completely dissolved in the E2-medium at 30 μM. Thus, an exact concentration cannot be stated for these compounds. Nevertheless, both compounds exerted no toxicity in the observation range (<30 μM).