Rational Design of a Circularly Permuted Flavin-Based Fluorescent Protein

Abstract

Flavin-based fluorescent proteins are oxygen-independent reporters that hold great promise for imaging anaerobic and hypoxic biological systems. In this study, we explored the feasibility of applying circular permutation, a valuable method for the creation of fluorescent sensors, to flavin-based fluorescent proteins. We used rational design and structural data to identify a suitable location for circular permutation in iLOV, a flavin-based reporter derived from A. thaliana. However, relocating the N- and C-termini to this position resulted in a significant reduction in fluorescence. This loss of fluorescence was reversible, however, by fusing dimerizing coiled coils at the new N- and C-termini to compensate for the increase in local chain entropy. Additionally, by inserting protease cleavage sites in circularly permuted iLOV, we developed two protease sensors and demonstrated their application in mammalian cells. In summary, our work establishes the first approach to engineer circularly permuted FbFPs optimized for high fluorescence and further showcases the utility of circularly permuted FbFPs to serve as a scaffold for sensor engineering.

Keywords: biosensing; circular permutation; flavin-based fluorescent proteins; light-oxygen-voltage sensing domain; reporter gene.

Figure 1:. Circular permutation of iLOV.

(a) Topology diagram showing secondary structure elements of iLOV with thick arrows and wavy lines representing β-sheets and α-helices respectively, and the intervening lines indicating loops. The β-sheets and α-helices are numbered alphabetically. The orange arrow indicates the site of circular permutation in the loop connecting the H and I β-sheets. (b) Whole-cell fluorescence of E. coli expressing iLOV constructs harboring insertions of the estrogen receptor ligand-binding domain in various loops. (c) Crystal structure of iLOV (4EES) showing location of circular permutation in the Hβ-Iβ loop. The structure is colored as per B-factors. (d) Whole-cell fluorescence of E. coli cells expressing iLOV constructs that are circularly permuted at the Gly95-Glu96 junction and have the original ends connected by a Jα-based linker. (e) Topology diagram showing secondary structure elements of circularly permuted iLOV with new termini and Glu96 and Gly95 and the Jα-linker connecting the original termini. Error bars represent the s.e.m. (n = 3). ** is P-value < 0.01; *** is P-value < 0.001.

Figure 2:. Optimization of circularly permuted iLOV.

(a) Fluorescence distribution of CHO cells transduced to express circularly permuted iLOV (cp). Fluorescence of cells transduced with a similar control vector (cntl) that lacks iLOV expression is shown for comparison. (b) Topology diagram showing secondary structure elements of circularly permuted iLOV with thick arrows and wavy lines representing β-sheets and α-helices respectively, and the intervening lines indicating loops. The coiled coil domains fused at the N- and C-termini are depicted as green wavy lines. (c) Fluorescence distribution of CHO cells transduced to express circularly permuted iLOV harboring E3/K3 coiled coils at the N and C termini. (d) Fluorescence distribution of CHO cells transduced to express circularly permuted iLOV harboring leucine zippers (Lz) at the N- and C-termini. (e) Ratio of mean fluorescence of CHO cells expressing circularly permuted iLOV constructs to the (auto)fluorescence of control cells lacking iLOV expression. (f) Representative images of CHO cells expressing circularly permuted iLOV (without coiled coils) or circularly permuted iLOV harboring leucine zippers, imaged in the green and red channels. Scale bar is 10 μm. (g) Fluorescence distribution of CHO cells transduced to express native (i.e., non-permuted) iLOV and iLOV harboring leucine zippers at the N- and C-termini. Error bars represent the s.d. (n = 4).

Figure 3:. Biochemical characterization of purified iLOV constructs.

(a) Denaturing polyacrylamide gel electrophoresis of purified iLOV reporters, including iLOV (lane 1), circularly permuted (cp) iLOV (lane 2), circularly permuted iLOV harboring leucine zippers (Lz) at each terminus (lane 3), and iLOV harboring leucine zippers at each terminus (lane 4). (b) Fluorescence excitation (dashed lines) and emission (solid lines) spectra of various iLOV-based reporters. (c) Thermal shift analysis depicting melting curves of various iLOV-based reporters. The melting curves were generated by incubating each protein at a given temperature (20–100 °C range) for 5 min and measuring the decrease in fluorescence emission (due to protein unfolding and/or loss of bound flavin).

Figure 4:. Engineering protease-sensing in circularly permuted iLOV.

(a) Topology diagram showing secondary structure elements of circularly permuted iLOV harboring leucine zippers (Lz) at the N- and C-termini and a TEVp cleavage site immediately after the Jα linker. (b) Fluorescence distribution of CHO cells transduced to express doxycycline-inducible TEVp and circularly permuted iLOV with a TEVp cleavage site, in the presence and absence of TEVp expression. Autofluorescence of cells transduced with a control vector (blnk) that lacks iLOV expression is shown for comparison. (c) Fluorescence distribution of CHO cells transduced to express doxycycline-inducible Mpro and circularly permuted iLOV with an Mpro cleavage site, in the presence and absence of Mpro expression. Autofluorescence of cells transduced with a control vector (blnk) that lacks iLOV expression is also shown for comparison. (d) Percent change in mean fluorescence of cells (−ΔF/F_o) expressing iLOV (non-permuted), TEVp and Mpro-sensing constructs, induced by expression of the respective protease. Error bars represent s.d. (n = 3).