Protein biochemistry: Evolving a better-expressing GPCR

signaling update home

SOURCE:

A directed evolution approach has been successful in designing G protein–coupled receptors (GPCRs) with greater stability and enhanced expression.

Sixty percent of all drugs target the class of membrane proteins known as G protein–coupled receptors (GPCRs), but the disconnect between their biomedical importance and the number of atomic resolution structures is very large, with only a handful of GPCR structures solved. This certainly does not reflect a lack of interest or attempts, but the tremendous challenges involved in GPCR expression, purification and crystallization. Even when overexpressed in heterologous systems, GPCRs typically express at very low levels in the cell membrane. They are also not very stable in detergents and exhibit conformational flexibility, making them difficult to crystallize. All of these bottlenecks add up, making the structure determination of GPCRs a Herculean task.

Andreas Plückthun of the University of Zurich and his colleagues hope to change this with a new directed evolution method intended to address these bottlenecks. According to Plückthun, the GPCR structure field has so far relied on "finding the lucky break," or basically using brute-force methods to obtain a structure that can in turn be used to model the rest of the family. However, "I always thought that all GPCRs are interesting and we have to find a method that eventually will make all of them amenable for study," he says.

Using the GPCR rat neurotensin receptor-1 (NTR1) as an example, the researchers tested whether they could modify its sequence via directed evolution to make the protein more expressible while still maintaining its function. Such an approach has not been tried before for GPCRs. They constructed a Ntsr1 library via error-prone PCR and expressed the constructs with N-terminal maltose binding protein and C-terminal thioredoxin fusion partners in Escherichia coli. Because GPCRs bind specific ligands, the researchers took advantage of high-throughput fluorescence-activated cell sorting (FACS) to identify cells expressing high levels of NTR1 variants that bound a fluorescently labeled ligand. After several rounds of directed evolution and FACS, they sequenced and analyzed the enriched clones.

They identified a mutant, named D03, which in the E. coli membrane exhibited a tenfold increase in expression compared to the wild-type NTR1. This mutant had just 14 nucleotide substitutions throughout the sequences encoding helices and loops, five of which were silent. Agonist binding to D03 was just as strong as to the wild-type NTR1, and the mutant maintained signaling properties when expressed in mammalian cells. Notes Plückthun: "The interesting finding, which to me at least was somewhat unexpected, is that the mutations showed improvement [in expression levels] in every expression system" that they tested, which also included the yeast Pichia pastoris. This suggests that the mutations introduced into the D03 mutant likely confer an overall stability to the protein, resulting in more robust expression. They also purified sixfold more D03 than the wild-type NTR1 from E. coli, and the mutant was more thermally stable in detergent-solubilized form, which may promote its crystallization.

Although the researchers have so far only reported results for NTR1, they are currently working on additional mutagenesis of NTR1 as well as testing the generality of the method for other GPCRs. Plückthun notes that the FACS-based selection method is likely to be applicable for evolving better-expressing variants of any membrane receptor that can bind a fluorescent ligand.

They also have yet to test whether their method can streamline the bottlenecks in GPCR crystallization. If such an approach does turn out to be general for evolving more crystallizable variants, it could be extremely powerful and have a major impact on our understanding of GPCR biology. "The interesting part is to understand how the ligand binds, the exact atomic details, and what the differences between an agonist and an antagonist are," says Plückthun. "I just don't think it can really be extrapolated from one model. We have to have an experimental access to basically the whole family."

Allison Doerr

References

Sarkar, C.A. et al. Directed evolution of a G protein–coupled receptor for expression, stability, and binding selectivity. Proc. Natl. Acad. Sci. USA 105, 14808–14813 (2008). | Article | PubMed |