BioCentury
ARTICLE | Distillery Techniques

Assays and screens

October 25, 2018 6:03 PM UTC

A bioinformatics screen based on transcriptomic data sets could help identify new targets for approved drugs. The screen involved 3 steps: analyzing 2,052 genes from eight signaling pathways in data sets from patients with various cancer types to identify gene signatures shared across multiple tumors; transcriptomic analysis of human cell lines treated with approved drugs to identify gene signatures differentially expressed in response to drug treatment; and computational comparison of each of the gene mutation signatures with the drug treatment signatures to predict if the drugs inhibit, activate or have no effect on the protein products of the mutated genes. In a proof-of-concept experiment, the screen predicted calcium channel blockers would also activate AMP-activated protein kinase (AMPK) or inhibit protein kinase B (AKT; AKT1; PKB; PKBA), and in 40 obese patients, those receiving calcium channel blockers had lower levels of bilirubin and higher levels of alkaline phosphatase activity in blood -- markers of AMPK activation and AKT inhibition -- than patients receiving other classes of antihypertensive drugs. In Caenorhabditis elegans, six drugs predicted by the screen to activate AMPK or inhibit AKT, including cyproheptadine, perhexiline and etilefrine, increased survival compared with no treatment. Next steps could include using the screen to identify targets for other classes of drugs.

The generic antihistamine cyproheptadine is marketed to treat allergies...