Biophysical methods, such as NMR and X-ray crystallography, are becoming increasingly popular early in lead generation, owing to their effectiveness in identifying low-affinity, low-molecular-mass 'fragments' and aiding their development into potent compounds. Lesuisse et al. now highlight the potential of such approaches by using X-ray-based screening of molecular fragments to develop nanomolar inhibitors of the SH2 domain of SRC, a kinase involved in bone resorption that represents a potential target for the treatment of osteoporosis.

The SH2 domain of SRC, which is involved in protein–protein interactions, seems to be crucial to its bone-resorbing activity, and this has therefore been the target of previous drug discovery efforts. However, although potent inhibitors that mimic the natural phosphate-containing peptide sequence that is recognized by the SRC SH2 domain — pYEEI — have been reported, the presence of the phosphotyrosine group that is necessary for strong binding in these cases is highly undesirable for drug candidates, as this group is readily hydrolysed, and has low membrane permeability.

The authors' search for compounds without these undesirable properties was prompted by two observations: first, that the binding of phenylphosphate (a fragment of phosphotyrosine) could be detected using conventional binding-assay techniques; and second, that crystals of SRC could be obtained that had the appropriate nature and quality to allow fragments to be soaked into them, such that the resultant structure could be determined by X-ray crystallography.

Two types of fragment — more than 150 in total — were screened both for binding affinity and by soaking them into SRC crystals to establish their binding mode: phenylphosphate-related fragments to elucidate optimal modification of the phenyl ring, and fragments that contained carboxylic acids, which have been reported as phosphate mimics. In the case of the acids, although no binding could be detected in the conventional assays, analysis of the crystal structures revealed that some of the fragments were forming interactions analogous to those involved in the binding of the natural ligand.

Incorporating the best phosphotyrosine mimetics indicated by these studies into the previously discovered potent inhibitors that mimicked the remainder of the pYEEI sequence resulted in several compounds with low nanomolar affinity for the SH2 domain, which could prove valuable in the cellular validation of the concept of SH2-domain inhibition.