Genetically Encoded RNA-Based Bioluminescence Resonance Energy Transfer (BRET) Sensors

Abstract

RNA-based nanostructures and molecular devices have become popular for developing biosensors and genetic regulators. These programmable RNA nanodevices can be genetically encoded and modularly engineered to detect various cellular targets and then induce output signals, most often a fluorescence readout. Although powerful, the high reliance of fluorescence on the external excitation light raises concerns about its high background, photobleaching, and phototoxicity. Bioluminescence signals can be an ideal complementary readout for these genetically encoded RNA nanodevices. However, RNA-based real-time bioluminescent reporters have been rarely developed. In this study, we reported the first type of genetically encoded RNA-based bioluminescence resonance energy transfer (BRET) sensors that can be used for real-time target detection in living cells. By coupling a luciferase bioluminescence donor with a fluorogenic RNA-based acceptor, our BRET system can be modularly designed to image and detect various cellular analytes. We expect that this novel RNA-based bioluminescent system can be potentially used broadly in bioanalysis and nanomedicine for engineering biosensors, characterizing cellular RNA-protein interactions, and high-throughput screening or in vivo imaging.

Keywords: bioluminescence resonance energy transfer; bioluminescent sensors; fluorogenic RNA aptamers.

Figure 1.

(a) Schematic illustration of a modular BRET sensor system based on TAR–Tat interaction-induced energy transfer between NanoLuc/substrate complex and fluorogenic RNA/dye pairs. The target binding to the aptamer region induces the folding of fluorogenic RNA to bind the dye molecule. While it is worth mentioning that data shown in the rest of Figure 1 was measured using RNA constructs without containing a target-binding aptamer region. (b) The I₆₁₂/I₄₆₀ (acceptor/donor) ratiometric BRET signal was measured at 25°C in a solution containing 100 nM t-NLuc, 1 μM RNA, 5 μM HBC620, and 5 mM Mg²⁺ at 30 s after adding 1% (v/v) furimazine substrate. S5–S10 represent the t-NLuc/TAR-Pepper construct containing 5–10-base-pair-long linkers, respectively. The mixture of t-NLuc with TAR, Pepper, and their simple mixture (TAR/Pepper) acted as the negative controls. (c, d) The BRET efficiency of Broccoli- and Mango II-based t-NLuc/RNA construct with different linker lengths (named as B1, B2, B3 and Ma1, Ma2, Ma3). The I₅₀₇/I₄₆₀ and I₅₃₅/I₄₆₀ ratiometric signals were measured in a solution containing 100 nM t-NLuc and 1 μM RNA at 25°C, in the presence of 20 μM DFHBI-1T or 1 μM TO1-biotin. The mixture of t-NLuc with TAR (without Broccoli or Mango II), Broccoli, Mango II (without TAR conjugation), or simple mixtures of Broccoli/Mango II and TAR (i.e., TAR/Broccoli and TAR/Mango) were used as negative controls. (e) The normalized I₆₁₂/I₄₆₀ BRET signal of the t-NLuc/S6 construct after mixing 100 nM t-NLuc with different concentrations of the S6 RNA. Shown are the mean and standard deviation (SD) values from three independent replicates. *p< 0.05, **p< 0.01, ***p< 0.001 in two-tailed student’s t-test.

Figure 2.

In vitro characterization of different RNA-based BRET sensors. The I₆₁₂/I₄₆₀ ratiometric signals of the optimal (a) t-NLuc/T1 tetracycline sensor, (d) t-NLuc/M1 SAM sensor and (g) t-NLuc/P3 ppGpp sensor were measured in the presence or absence of 100 μM tetracycline (Tet), SAM or ppGpp. The mixture of t-NLuc with TAR or S6 acted as a negative control and positive control, respectively. (b, e, h) Dose response curve of the t-NLuc/T1 tetracycline sensor, t-NLuc/M1 SAM sensor and t-NLuc/P3 ppGpp sensor after adding different amounts of tetracycline, SAM or ppGpp. (c, f, i) Selectivity of the t-NLuc/T1 tetracycline sensor, t-NLuc/M1 SAM sensor and t-NLuc/P3 ppGpp sensor was measured in a solution containing 100 μM counter ligands. Control was measured without adding ligands. All these measurements were performed in a solution containing 100 nM t-NLuc, 1 μM RNA, and 5 μM HBC620 at 25°C after adding 1% (v/v) furimazine substrate. Shown are the mean and the standard error of the mean values from three independent replicates in each case.

Figure 3.

Target detection in living BL21 Star (DE3) E. coli cells with different RNA-based BRET sensors. The ratiometric signals of t-NLuc/T1 sensor-expressing cells in the presence of different amounts of tetracycline using (a) plate reader (I₆₁₂/I₄₆₀) and (d) IVIS SpectrumCT (I₆₂₀/I₅₀₀). The ratiometric signals of t-NLuc/M1 sensor-expressing cells in the presence of different amounts of L-methionine, a precursor for the cellular synthesis of SAM, using (b) plate reader (I₆₁₂/I₄₆₀) and (e) IVIS SpectrumCT (I₆₂₀/I₅₀₀). The ratiometric signals of t-NLuc/P3 sensor-expressing cells in the presence of different amounts of serine hydroxamate (SHX) for inducing bacterial starvation and ppGpp generation, using (c) plate reader (I₆₁₂/I₄₆₀) and (f) IVIS SpectrumCT (I₆₂₀/I₅₀₀). Cells expressing t-NLuc/TAR or t-NLuc/S6 acted as a negative control and positive control, respectively. The measurement was performed at 25°C in the presence of 5 μM HBC620 right after adding 1% (v/v) furimazine substrate. Shown are the mean and SD values from three independent replicates. *p< 0.05, ***p< 0.001 in two-tailed student’s t-test; ns, not significant.

Figure 4.

(a) I₆₁₂/I₄₆₀ ratiometric BRET signals of genetically encoded t-NLuc/T1 sensors in BL21 Star (DE3) cell lysates in the absence or presence of additional 100 μM tetracycline. (b) I₆₁₂/I₄₆₀ ratiometric BRET signals of genetically encoded t-NLuc/M1 sensors in BL21 Star (DE3) cell lysates in the absence or presence of additional 1 mM SAM. (c) I₆₁₂/I₄₆₀ ratiometric BRET signals of genetically encoded t-NLuc/P3 sensors in BL21 Star (DE3) cell lysates in the absence or presence of additional 100 μM ppGpp. t-NLuc/TAR (without the Pepper unit) acted as a negative control, and the t-NLuc/S6 construct acted as a positive control. The measurement was performed at 25°C in the presence of 5 μM HBC620 right after adding 1% (v/v) furimazine substrate. (d, e) Bright field and fluorescence channel imaging of HEK-293T cells at 24 h after transfection with (d) mini-CMV-t-NLuc and pAV-U6+27-Tornado-TAR or with (e) mini-CMV-t-NLuc and pAV-U6+27-Tornado-S6. Fluorescence images were collected through a 593/20 nm filter after irradiation with a 561 nm laser at 25°C in the presence of 5 μM HBC620. Scale bar, 20 μm. (f) The I₆₄₅/I₄₆₀ ratiometric signals of HEK-293T cells after confirming the successful transfection with mini-CMV-t-NLuc and pAV-U6+27-Tornado-TAR or pAV-U6+27-Tornado-S6. The measurement was performed in a plate reader at 25°C in the presence of 5 μM HBC620 at 30 s after adding 1% (v/v) furimazine substrate. Shown are the mean and SD values from three independent replicates. *p< 0.05, ***p< 0.001 in two-tailed student’s t-test.