A Hit In Silico

Computer-assisted design has made possible the synthesis of a reverse transcriptase inhibitor with unprecedented potency against HIV-1 activity in human T cells (J. Med. Chem., DOI: 10.1021/jm201134m). The work helps validate computational design as a route to promising drug candidates.

Chemistry professor William Jorgensen and pharmacology professor Karen S. Anderson, both at Yale University, and their colleagues set out to develop a nonnucleoside reverse transcriptase inhibitor (NNRTI). This class of inhibitors binds in a side pocket of HIV’s reverse transcriptase enzyme, inhibiting motions critical in forming infective viral DNA from the virus’s genomic RNA. The group used computational design to tweak an original computer-docking hit.

In vitro tests show that the compound, called JLJ494, can inhibit replication of HIV-1 in picomolar-range doses, an order of magnitude better potency than other NNRTIs. It hasn’t yet been tested in animals. Yale has filed for a patent on the compound class.

JLJ494’s potential use as a drug may be limited because NNRTIs lose effectiveness quickly when HIV-1 mutates, notes Dirk Jochmans of Catholic University of Leuven, Belgium. However, the computer design was “very successful,” he says.

Charles W. Flexner of Johns Hopkins University notes that five NNRTIs are already used in the U.S. and Europe. But a stronger compound requiring small doses would be easier to manufacture as well as to distribute and administer to people in poor countries.

Jorgensen, Anderson, and coworkers screened more than 2 million compounds using Schrödinger’s Glide docking software. They selected nine promising candidates, three of which showed anti-HIV activity. The group then used further computations to optimize the structure of what became JLJ494, a catechol diether derivative.

They suspect that JLJ494’s potency may stem from its ability to interact with a proline residue on reverse transcriptase. If true, researchers could then optimize this interaction in future designs.

“I think the striking thing is that computations led us to compounds that are the most potent reported in this class,” Jorgensen says.