DNA Display Selection of Peptide Ligands for a Full-Length Human G Protein-Coupled Receptor on CHO-K1 Cells

et al. (2012) DNA Display Selection of Peptide Ligands for a Full-Length Human G Protein- Coupled Receptor on CHO-K1 Cells. PLoS ONE 7(1): e30084. doi:10.1371/journal.pone.0030084 DNA Display Selection of Peptide Ligands for a Full- Length Human G Protein-Coupled Receptor on CHO-K1 Cells Nobutaka Matsumura 0 Kenichi Horisawa 0 Hiroshi Yanagawa 0 Maria Gasset, Consejo Superior de Investigaciones Cientificas, Spain 0 Department of Biosciences and Informatics, Keio University , Yokohama , Japan The G protein-coupled receptors (GPCRs), which form the largest group of transmembrane proteins involved in signal transduction, are major targets of currently available drugs. Thus, the search for cognate and surrogate peptide ligands for GPCRs is of both basic and therapeutic interest. Here we describe the application of an in vitro DNA display technology to screening libraries of peptide ligands for full-length GPCRs expressed on whole cells. We used human angiotensin II (Ang II) type-1 receptor (hAT1R) as a model GPCR. Under improved selection conditions using hAT1R-expressing Chinese hamster ovary (CHO)-K1 cells as bait, we confirmed that Ang II gene could be enriched more than 10,000-fold after four rounds of selection. Further, we successfully selected diverse Ang II-like peptides from randomized peptide libraries. The results provide more precise information on the sequence-function relationships of hAT1R ligands than can be obtained by conventional alanine-scanning mutagenesis. Completely in vitro DNA display can overcome the limitations of current display technologies and is expected to prove widely useful for screening diverse libraries of mutant peptide and protein ligands for receptors that can be expressed functionally on the surface of CHO-K1 cells. - Funding: This work was supported by the Research for Promoting Technological Seeds grant (09-479) from the Japan Science and Technology Agency; the Industrial Technology Research Grant Program (03A01007a) and a Grant for Practical Application of University Research & Development Results under the Matching Fund Method from the New Energy and Industrial Technology Development Organization of Japan; a Grant-in-Aid for Scientific Research (19360377) from the Japan Society for the Promotion of Science; and a Special Coordination Fund grant from the Ministry of Education, Culture, Sports, Science and Technology of Japan. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. The superfamily of G protein-coupled receptors (GPCRs) [1] is the largest and most diverse group of cell-surface proteins involved in signal transmission. Although a large number of GPCRs has been identified in the human genome project [2,3], more than 100 of them have no known physiologically relevant ligand yet [4,5], and they are classified as orphan GPCRs. Since GPCRs are major targets for todays drugs [6], the search for cognate and surrogate peptide ligands for GPCRs is of both basic and therapeutic interest [7,8]. Conventional analysis of the specificity of the interaction between GPCRs and peptide ligands involves the mutation of individual amino acids by peptide synthetic methods (e.g., Alascanning), followed by measurement of binding affinity or receptor activation. However, the sequence space that can be searched with this standard strategy is quite limited. As a powerful alternative strategy, phage display has been used to screen peptides that bind to GPCRs expressed on mammalian cells [913], but the library sizes and the sequence varieties in a phage library are limited by the transformation efficiency and biological constraints of the host bacteria. This limitation can potentially be overcome by using totally in vitro selection systems, such as ribosome display and mRNA display, which employ cell-free protein synthesis [1417]. Recently, mRNA display was used to screen peptide ligands that bind to the N-terminal extracellular domain of a class B GPCR immobilized on beads [18], but such an RNA-tagging method requires strictly RNase-free conditions and cannot easily be applied to selection targeting full-length GPCRs expressed on the cell surface. We have previously developed a DNA display system called STABLE (STreptAvidin-Biotin Linkage in Emulsions) [1921], in which streptavidin-fused peptides are linked with their encoding DNA via biotin labels in a cell-free transcription/translation system compartmentalized in water-in-oil emulsions. This method allows completely in vitro selection of a stable DNA-tagged peptide library with large diversity in the presence of RNase. In this study, we applied the DNA display system to in vitro selection of peptide ligands for a full-length GPCR expressed on whole cells. As a model to test our screening strategy, we used a well-known GPCR, human angiotensin II (Ang II) type-1 receptor (hAT1R), which is significantly involved in cardiovascular diseases. Under improved selection conditions using hAT1R-expressing mammalian cells as bait, the Ang II gene was enriched from model libraries (1:100 or 1:10,000 mixture of streptavidin-fused Ang II and streptavidin genes). Further, various Ang II-like peptides were successfully selected from randomized peptide libraries, and their binding activity and biological function were characterized to elucidate the sequence-function relationship of hAT1R ligands. selected DNA from DNA-peptide conjugates bound to GPCRs by means of simple photocleavage [22]. Strategy for in vitro selection of GPCR-ligands We improved and applied the STABLE DNA display system [1922] for in vitro selection of GPCR-ligands on whole cells (Fig. 1). In this system, the linkage of DNA (genotype) and peptide (phenotype) was accomplished in water-in-oil emulsions containing an in vitro transcription/translation system, in which one DNA molecule was caged in each reversed micelle on average [23]. A stable binding of streptavidin with biotin was used as the connector between DNA and its translated products [19]. The number of DNA-peptide conjugates in a library (i.e., library size) is comparable with the number of emulsion droplets (1091010 per 1 ml of emulsion). The DNA-displayed peptide library was incubated with GPCR-expressing cells in the presence of a GRGDS pentapeptide to inhibit undesired binding of an RGDlike sequence within streptavidin to integrins on the cell surfaces [24]. We also added 0.5 M sucrose to the binding buffer to inhibit internalization of agonist peptides by receptor-mediated endocytosis [25]. Furthermore, in order to repress the large background of cell-surface proteins, glycans and lipids, the library was preincubated with Mock cells without recombinant GPCR to remove nonspecific binders before incubation with the GPCRexpressing cells (not shown in Fig. 1). A Chinese hamster ovary cell line CHO-K1 is suited for this p (...truncated)