Design and Prototyping of Genetically Encoded Arsenic Biosensors Based on Transcriptional Regulator AfArsR

Abstract

Genetically encoded biosensors based on engineered fluorescent proteins (FPs) are essential tools for monitoring the dynamics of specific ions and molecules in biological systems. Arsenic ion in the +3 oxidation state (As³⁺) is highly toxic to cells due to its ability to bind to protein thiol groups, leading to inhibition of protein function, disruption of protein-protein interactions, and eventually to cell death. A genetically encoded biosensor for the detection of As³⁺ could potentially facilitate the investigation of such toxicity both in vitro and in vivo. Here, we designed and developed two prototype genetically encoded arsenic biosensors (GEARs), based on a bacterial As³⁺ responsive transcriptional factor AfArsR from Acidithiobacillus ferrooxidans. We constructed FRET-based GEAR biosensors by insertion of AfArsR between FP acceptor/donor FRET pairs. We further designed and engineered single FP-based GEAR biosensors by insertion of AfArsR into GFP. These constructs represent prototypes for a new family of biosensors based on the ArsR transcriptional factor scaffold. Further improvements of the GEAR biosensor family could lead to variants with suitable performance for detection of As³⁺ in various biological and environmental systems.

Keywords: FRET and FP-based arsenic biosensors; arsenic biosensors (GEARs); genetically encoded biosensor.

Figure 1

Design and characterization of the FRET-based As³⁺ biosensor prototype GEAR-CV1. (A) Design and gene structure of the genetically encoded As³⁺ indicator GEAR-CV1. (B) Schematic representation of expected FRET sensing mechanism of GEAR-CV1. (C) GEAR-CV1 normalized emission fluorescence spectrum (excitation at 430 nm emission scanned from 450 nm to 600 nm) in Tris-buffered saline (TBS) with (red) and without (blue) As³⁺. (D) FRET acceptor-to-donor fluorescence ratio (R = F₅₃₀/F₄₇₅) of purified GEAR-CV1 protein when titrated with As³⁺.

Figure 2

Mutational study of cysteine residues in the AfArsR arsenic binding site using GEAR-CV1. (A) The structure of AfArsR (PDB ID: 6J05) with cysteine residues 95, 96, and 102 (sticks) binding As³⁺ (purple sphere). FRET ratios of cysteine-mutated variants based on GEAR-CV1, in Tris-buffered saline without (blue) and with (red) 1 mM As³⁺, including single-mutants Cys95Ala (B) Cys96Ala (C), and Cys102Ala (D) double-mutant Cys95Ala/Cys96Ala (E), and triple-mutant Cys95Ala/Cys96Ala/Cys102Ala (F). Significant differences between pairs are indicated as **** (p < 0.0001), n.s. not significant (p > 0.05).

Figure 3

Optimization of GEAR-CV1 with structure-guided deletion and FRET donor/acceptor alteration. (A) Deletion of AfArsR C-terminus residue resulting in GEAR-CV2 with improved FRET ratio change. (B) FRET ratio of GEAR-CV1, in Tris-buffered saline without (blue) and with (red) 1 mM As³⁺. (C) FRET ratio of GEAR-CV2, in Tris-buffered saline without (blue) and with (red) 1 mM As³⁺. (D–F) FRET-based GEAR variants with alternative FP donor and acceptor. FRET ratios of GEAR-TV1 (mTFP1/cpVenus) (D), GEAR-CC1 (mCerulean3/mCitrine) (E), and GEAR-TC1 (mTFP1/mCitrine) (F), in Tris-buffered saline without (blue) and with (red) 1 mM As³⁺. Significant differences between pairs are indicated as **** (p < 0.0001), ** (p < 0.01), and * (p < 0.05).

Figure 4

Design and engineering of single FP-based As³⁺ biosensor prototypes GEAR-G1 and GEAR-G2. (A) Design and DNA construction of the genetically encoded As³⁺ indicator GEAR-G1. (B) Schematic representation of the sensing mechanism of GEAR-G1. (C) GEAR-G1 normalized excitation and emission fluorescence spectrum in Tris-buffered saline (TBS) with (red) and without (blue) As³⁺. (D) Rational optimization of residues 145 and 148 of GFP (blue and magenta residues), the linker regions (grey residues) for improved single FP-based As³⁺ biosensor prototype GEAR-G2. (E) GEAR-G2 normalized excitation and emission fluorescence spectrum in Tris-buffered saline (TBS) with (red) and without (blue) As³⁺.

Figure 5

(A) Structure of GEAR G1 predicted by Robetta. (B) Snapshots of structure of GEAR-G1 at different time scales of MD simulation. Yellow, purple, blue, and cyan/white colors depict various secondary structure elements such as β-sheets, α-helix, short α-helices, and coil/turns, respectively. (C) Dictionary of Secondary Structure of GEAR-G1 protein (DSSP), as obtained from 150 ns of MD simulation. (D) RMSD, Rg, and RMSF plots as calculated from the α-carbons of the GEAR-G1 protein. The RMSF plot has been generated for the last 100 ns of the trajectory. Residues 146–263 correspond to AfArsR, and residues 1–145 and 264–353 correspond to GFP. (E) Dynamic cross correlation matrix map of the trajectory for GEAR-G1 protein complex. The value ranges from +1 (cyan color) to −1 (purple color). Positive values represent positive inter-residual correlation, while negative values represent negative inter-residual correlation. The map depicts that both a positive and negative inter-residual correlation exists in the atoms of GEAR-G1 protein, which is essential for its function.

Figure 6

(A) Depiction of the proportion of variance (%) against its Eigenvalue rank. (B) Visualization of dynamics of PC1 for depiction of the fluctuating regions. (C–E) Projection of the trajectory formed by the first three principal component analyses based on k-means values.