[74] (Figure 1.11), utilizing in situ generated singlet oxygen for generation of a peroxide, followed by Hock cleavage. While this flow chemistry process was not implemented in the ultimate production process, which involves photooxidation of activated dihydroartemisinic acid and is capable of delivering average quantities of 370 kg per batch [75], the methodology still has great potential for future applications. In particular, it could be demonstrated on an impressive example that facile photochemistry can be applied in process chemistry by rethinking the production process and converting the synthesis from large‐scale bulk synthesis to steady small‐scale synthesis by using a continuous process, which also is capable of delivering large quantities of drug substance.
1.4.4 Carfilzomib
The discovery of carfilzomib started from a regular literature search. Craig Crews (Yale University) was searching for new project ideas and reviewed past issues of the Journal of Antibiotics. He came across a compound called epoxomicin (Figure 1.12) [76] that caught his attention. The compound was isolated from an Actinomycete strain and displayed interesting cytotoxic activity against various cancer cell lines, as well as activity in an in vivo B16 leukemia model. The stereochemistry was not determined, but it had an exposed epoxide as a rather unusual structural feature. The molecule was actually discovered by Japanese researchers at BMS, but as the mechanism of action was unknown and the drug‐like properties of the compound were rather poor, BMS decided to drop the project. Crews did not intend to start a drug discovery project, but was rather interested in applying emerging chemical biology techniques to the molecule and unravel its mode of action. He completed the first total synthesis of the molecule, which also allowed determination of the previously unknown stereochemistry [77]. By employing a biotinylated derivative, he determined that epoxomicin specifically targets the proteasome [78]. The proteasome, a protein complex of about 1700 kDa, is responsible for degradation of misfolded proteins and exists in all eukaryotic cells and archaea, as well as in several prokaryotes. Shutting down the proteasome will significantly disturb normal cellular processes and will quickly kill the cell. However, cells with high replication rates should be more dependent on optimal functionality of the proteasome, thus opening an opportunity for cancer therapy. Several proteasome inhibitors, natural and synthetic, had already been reported in the literature [79], but their structures usually contained very reactive warheads, resulting in insufficient compound selectivity. Crews showed that epoxomicin, in contrast to other compounds with reactive warheads, selectively inhibited the proteasome. He started a collaboration with German Nobel laureate Robert Huber (Max Planck Institute for Biochemistry, in Martinsried, Germany) and solved the crystal structure of epoxomicin bound to the proteasome [80]. They discovered that epoxomicin reacted with the N‐terminal threonine moiety of the 20S proteasome (Figure 1.12), forming a stable morpholino ring through successive epoxide opening by the terminal amino group and attack of the nucleophilic hydroxyl group on the ketone of epoxomicin. This specificity was remarkable, which prompted Crews and his coworkers to derivatize and improve epoxomicin.
Figure 1.12 Epoxomicin binding to the 20S‐ribosome.
Source: Based on Hanada et al. [76].
After several rounds of optimization, first systematically varying the individual positions of the tetrapeptide and then combining the optimized residues in one molecule, they came up with a compound they later termed YU‐101 (Figure 1.13) [81]. The compound did show significantly enhanced activity compared with epoxomicin and PS‐341 (bortezomib), a dipeptidyl boronic acid derivative of epoxomicin developed by a biotech company called ProScript. Bortezomib was later acquired by Millennium Pharmaceuticals and became the FDA approved medication Velcade™, used for treatment for treatment of multiple myeloma and mantle cell lymphoma. YU‐101 was licensed to Proteolix, a startup company founded by Craig Crews and Raymond J. Deshaies (California Institute of Technology). Proteolix was dedicated to the discovery of drugs targeting the proteasome. Scientists at Proteolix continued optimization and finally selected carfilzomib for preclinical and later clinical development. Proteolix was acquired by Onyx in 2009 for a nominal value of US$ 810 million. Carfilzomib (Kyprolis™, Figure 1.13) was approved by the FDA in July 2012 for treatment of advanced multiple myeloma and in 2015/2016 in combination with dexamethasone or lenalidomide and dexamethasone for treatment of refractory melanoma.
Figure 1.13 From epoxomicin to carfilzomib.
1.5 Biologic Drugs
1.5.1 Insulin
Contributions to drug discovery from academic groups are not limited to small molecules. Biologic drugs bear tremendous promise, and significant progress has been made to use them as therapeutics. The earliest reported example is the discovery of insulin [82]. The optimization of insulin to adapt short‐ and long‐acting profiles has been described earlier in this series [83]. Interestingly, this work, albeit carried out in the laboratories of renowned pharmacologist John Macleod, was started by a rather inexperienced student, Frederick Grant Banting.
After returning from World War I, Banting worked as a lecturer in fall of 1920. Preparing for a lecture on the pancreas, a recently published article caught his attention describing the observation of surviving islets in an obstructed, atrophic pancreas. He assumed that degrading enzymes could be responsible for losing the active principle of pancreatic secrete and that these enzymes would likely be produced in acinar cells. Thus he developed the idea that by ligation of the pancreas and induction of atrophy, it may be possible to selectively destroy acinar cells and thus deplete degrading enzymes while maintaining the active blood sugar‐lowering ingredient, which could then possibly be isolated by extraction. Enthusiastically he contacted Macleod, a proven authority in the field of diabetes. Macleod was skeptical that the approach could work, as many others had failed in isolating active pancreatic extracts before. He also easily noted that Banting only possessed textbook knowledge on diabetes, was not acquainted with recent literature on the topic, and also did not have the practical surgical experience to successfully perform the complicated procedures. However, after several meetings, Macleod agreed to offer him a (non‐paid) opportunity to experiment in his laboratories and asked one of his student assistants, Charles H. Best, to assist Banting in the proposed research. They started their experimental work on 17 May 1921, but it quickly turned out that Banting overestimated his surgical skills and the dogs faded quickly. However, Banting and Best subsequently formed an experienced team, and within 2.5 months they managed to treat a pancreatectomized dog with an extract isolated from excised pancreata of other dogs, which was able to transiently reduce its blood glucose. This result caused great excitement, but provision of extracts from duct‐ligated pancreata was a laborious method with very limited throughput. In August 1921 they developed the idea to utilize fetal calf pancreata, being available from butchers, as these contained less acinar cells (which are responsible for excreting digestive enzymes and thus would lead to destruction of the – yet to be discovered – insulin). This turned out to be successful. The process was further improved significantly when they decided to use alcohol for extraction of the fetal pancreata. The alcohol extract was significantly easier to concentrate than the previously used saline solution. On 11 December 1921, they decided to use the established protocol on an adult bovine pancreas and for the first time, this extract also displayed a strong glucose‐lowering
