RESEARCH HIGHLIGHTS

PROTEIN BIOCHEMISTRY Designer modulators

Through convergent engineering research- ers modified both the surface and a GDP small molecule, creating a specific interac- analog tion to manipulate protein activity. Small molecules can insert themselves into Wild-type GDP the most awkward places, inhibiting active H-Ras analog sites or regulatory regions to change pro- Mutant methods tein function. For years investigators have H-Ras GDP exploited these small molecules to manipu- analog late , but some proteins, including G proteins, have evaded such control. .com/nature e Members of the G-protein superfamily Figure 1 | Convergent engineering. Alteration of bind guanine nucleotides to set the protein the guanine nucleotide on H-Ras by

.natur conformation for activity: GTP makes G site-directed mutagenesis yielded a mutant w proteins active, whereas GDP turns them H-Ras that binds a nucleotide analog, allowing ‘off’. But the extremely strong affinity of G for specific regulation of the mutant protein. proteins for GDP and GTP has made isola- http://ww

tion of small-molecule modulators difficult. analog that fit the new binding pocket better Recently, however, Kavita Shah of Purdue than GDP did. oup r University and her colleagues showed that Using classic H-Ras activity assays, the G G proteins can be controlled after modifica- researchers showed that the new system of tion of both the small molecule and protein mutant protein and guanine nucleotide ana-

lishing through convergent engineering (Fig. 1). log could signal alongside wild-type H-Ras. b As molecular switches, G proteins are But the investigators could now specifically Pu important signaling factors, and are often activate mutant H-Ras to induce a persistent deregulated in and other diseases. ‘on’ conformation. In one experiment, Shah Presently available genetic techniques, how- and colleagues isolated complexes of mutant Nature

7 ever, do not allow G proteins to be manipu- H-Ras and its effectors, and even identified a lated during an experiment. In contrast, previously unknown H-Ras effector. 200

© “small-molecule modulators provide tempo- This ability to regulate H-Ras is likely ral control, which should greatly help in dis- transferable to the other ~200 known G pro- secting highly dynamic signaling cascades,” teins. Shah explains, “the residues mutated says Shah. in H-Ras are highly conserved across the To that end, Shah and colleagues designed G- , so three-dimensional an H-Ras protein that could be modulated by structural information is not required, and an orthogonal small molecule. They looked this approach should be widely applicable.” for mutations that would not affect the wild- To prove their point, the investigators made type function of H-Ras, in particular the the same mutations in another G protein, nucleotide-directed . Rap1B, and found the same responsiveness They hoped that substitution with alanines at to the same guanine nucleotide analogs. two positions in the guanine-ring binding site It is the strategy, however, that is particu- would permit an orthogonal small molecule larly exciting. By rationally engineering the with bulkier groups to fit into the binding site small molecule-protein interface, Shah and without disturbing the switch mechanism or colleagues generated new tools to study the substrate specificity. many roles of G proteins. She believes that To isolate small molecules that might this technique “can be a valid and valuable regulate the mutant H-Ras, the researchers approach to specifically modulate individual screened guanine nucleotide analogs. These members of protein superfamilies possessing molecules were a good starting material highly conserved active sites.” because, as Shah puts it, “orthogonal small Katherine Stevens molecules would likely use similar bind- ing modes, and therefore inactivate or acti- RESEARCH PAPERS Vincent, F. et al. Engineering unnatural nucleotide vate the mutant .” They identified a specificity to probe G protein signaling. Chem. Biol. doubly phosphorylated guanine nucleotide 14, 1007–1018 (2007).

988 | VOL.4 NO.12 | DECEMBER 2007 | NATURE METHODS