103(P-42) Realization of Skeletally Diverse Library from a Steroidal Compound Using Skeletal Transformation Strategy
Details
Our research goal is to make possible systematic dissection of biological events by use of small molecules as biological probes (chemical genetics). To discover bioactive compounds for diverse biological targets efficiently, we constructed an integrated platform performing compound synthesis, biological screening, and data store (Figure 1). Generating a high level of skeletal diversity in library compounds is important to get as many biologically active compounds as possible for diverse targets. Here, we proposed a new strategy called as skeletal transformation to get a skeletal diversity in a library. The key of the strategy is making better use of intramolecular reactions and the ideal order of the inter- and intramolecular reactions in the library synthesis. Intramolecular reactions after conventional intermolecular library synthesis give another skeletal compound accompanied by building blocks. Consequently, the strategy is efficient to generate building blocks diversity as well as skeletal diversity. We synthesized a skeletally diverse library from a steroidal compound 3 by the skeletal transformation strategy as shown in Scheme 3. For the first and second steps usual liquid phase conditions were applicable. We found Et_2AlCl could promote the Diels-Alder reaction at the third step without cleavage of silyl linker. Simple heating of beads was enough for the last retro-Diels-Alder reaction. We screened building blocks for each step, and selected 41 nucleophiles (6 thiols; 25 secondary amines; and 10 primary amines) for the first step, 15 reagents for the second step, and 12 ynones for the third step as shown in Figure 2. Theoretically the combination of these building blocks can produce a maximum of 4275 distinct compounds. We synthesized the library encoded with chemical tags using 25650 beads (4275 compounds×6 copy) as shown in Figure 3. The purity of compounds in library part 1, 2, and 3 were 78%, 71%, and 62% respectively. The intensity of tags was enough for decoding by CAN-cleavage and GC analysis. Biological screenings of the library synthesized here is currently in progress.
