[P005] Poster

12:10 PM - 1:00 PM JST(3:10 AM - 4:00 AM UTC)

[P03-443] A single fluorescent protein-based plant hormone biosensor employing an antibody developed through live-cell screening and computational simulations

○ZIXU FENG¹, Hiroshi Aihara¹, Shun Sakuraba^2,3, Yanyan Hou⁴, Keisuke Yoshida¹, Takanobu Yasuda¹, Bo Zhu¹, Toshiro Ito⁵, Tetsuya Kitaguchi^1,4 (1. Science Tokyo (Japan), 2. QST (Japan), 3. Chiba U (Japan), 4. WABIOS (Singapore), 5. NAIST (Japan))

Keywords:

Plant hormone,genetically encoded biosensor,fluorescent protein-based biosensor,antibody,live-cell screening,MD simulation

Plant hormones are small signaling molecules that regulate plant developmental and physiological processes. Especially, gibberellin (GA) controls seed germination and stem elongation and acts together with other plant hormones. However, the precise dynamics and cooperative relationships with these molecules remain to be fully understood.To understand GA dynamics in living cells, we aimed to develop a genetically encoded biosensor applicable to multi-color imaging. We first generated a prototype GA-Flashbody by inserting the circularly permutated GFP (cpGFP) between the variable fragments of heavy and light chains from anti-GA antibody¹, which is based on Flashbody design². The prototype was expressed in E. coli and tested in sonicated lysate, where GA induced only a small change in fluorescence intensity. Therefore, we extensively screened the two linkers between cpGFP and the antibody fragment, which were promising targets for improvement, using semi-rational design and saturation mutagenesis. We successfully yielded a variant with ~3-fold fluorescence response in vitro. We then expressed this variant in living E. coli and mammalian cells, which allow rapid and easy handling, and in protoplasts, whose preparation takes considerable time, to evaluate its function in living cells. The fluorescence response of the variant in E. coli and mammalian cells was limited (~1.1-fold), and the fluorescence in protoplasts was too dim to obtain the response. We speculated that the intracellular reducing environment interferes with disulfide bond formation in the antibody-derived domain, causing misfolding and loss of function of the prototype. Therefore, to improve the response in intracellular environment, we tried to identify the mutations by combining live-cell screening and computational simulations. In live-cell screening, we generated a library by error-prone PCR and screened the library in living E. coli using fluorescence-activated cell sorting (FACS) to enrich brighter variants, then performed colony screening on LB agar plates by spraying GA to identify clones exhibiting greater response. In computational simulations, we perform two molecular dynamic (MD) simulations followed by free energy calculations to identify mutations that stabilize structure, with the expectation that the mutants function in the intracellular environment. The 1st simulation evaluated scFv folding stability using published structure¹, while the 2nd did both folding and GA-binding stability using predicted full GA-Flashbody structure. Mutations with greater negative free energy than the wildtype, or with greater negative free energy in the GA-bound than unbound state, were selected for more stable structures in living cells.Finally, by integrating two mutations from the live-cell screening and three mutations from simulations, we found a variant with a 2.2-fold response in living E. coli, and named as GA-Flashbody. When expressed in mammalian cells, GA-Flashbody showed more than 3-fold fluorescence response, demonstrating a significant improved GA-dependent response under intracellular environment. We next plan to evaluate its performance in living plant cells, protoplasts. Future applications include generating transgenic plants to visualize GA dynamics in vivo, which will help clarify GA dynamics and its cooperative and hierarchical relationships with other hormones through multicolor imaging with high spatiotemporal resolution.