06 January 2010

Fragments in the Clinic: DG-051

Last August we highlighted work from deCODE on their leukotriene A4 hydrolase (LTA4H) program. That paper described the construction of a fragment library based on naturally occurring compounds, crystallographic screening against LTA4H, and optimization of inhibitors for this cardiovascular disease target. In a new paper published in J. Med. Chem., the researchers provide a fuller description of the discovery of the resulting clinical compound, DG-051.

As noted in the previous paper, crystallographic screening of deCODE’s fragment library identified several hydrophobic hits such as Compound 6 (see figure). At the same time, the researchers were aware of research from Searle that had produced inhibitors such as Compound 5. Appending the pyrrolidine of this compound onto deCODE’s fragment led to a modest increase in potency (Compound 9), though the resultant compound was still orders of magnitude weaker than Compound 5. Crystallography suggested a couple bad interactions in Compound 9 compared to Compound 5, so the researchers modified Compound 5 to generate Compound 14, which was active in a whole blood assay but suffered from rapid metabolism. Replacing the central methylene with an oxygen and adding a chlorine (Compound 17) improved biochemical potency slightly while dramatically improving pharmacokinetics.

Several crystal structures of LTA4H showed an acetate ion bound to the catalytic zinc, and the researchers sought to combine this “fragment” with their existing series, generating clinical compound DG-051. This did not lead to an improvement in biochemical potency (and actually decreased ligand efficiency), but it did lead to a roughly ten-fold improvement in potency in the whole-blood assay, as well as improvements in solubility and DMPK parameters. Interestingly, both enantiomers were equipotent, and the S-enantiomer was ultimately chosen due to ease of synthesis.

This story could be seen as an example of what has been called “fragment-assisted drug discovery:” unlike AT9283 or Indeglitazar, the fragments identified (acetate aside) didn’t end up in the clinical compound, and it could be argued that the initial lead was taken from the literature. But information gleaned studying the fragments fed into the design of a molecule that was sufficiently active, stable, selective, and novel for development.

The article states that DG-051 entered phase 2 clinical trials “for the prevention of myocardial infarction and stroke”, although no reports of development appear in clinicaltrials.gov. Also, in what has been an all-too-common event over the past year, deCODE filed for Chapter 11 bankruptcy and announced that it planned to sell “substantially all of its assets.” Practical Fragments wishes the best of luck to all the folks there. Happily the structural biology and fragment-screening group has (re)gained independence as Emerald BioStructures.

No comments: