Light‐Activated Gene Expression System Using a Caging‐Group‐Free Photoactivatable Dye

Abstract

Optical regulation of transcription using chemical compounds is an effective strategy to manipulate gene expression spatiotemporally. Conventional caging approaches with photoremovable protecting groups may require intense UV‐light exposure and release potentially toxic byproducts. To address these problems, here we developed a light‐mediated transcriptional regulation system by combining a caging‐group‐free photoactivatable dye PaX₅₆₀ and a multidrug‐binding transcriptional regulator QacR. The cationic dye generated from PaX₅₆₀ through traceless photoconversion bound QacR and reduced its repressor function, resulting in transcriptional activation. Importantly, this system allowed transcriptional activation with a large dynamic range under mild visible light exposure and simultaneous detection of the state of the photoactivated effector. This module was integrated into the T7 RNA polymerase expression system to demonstrate light‐activated transcription in vitro and in living cells.

A light‐induced gene expression system was developed using a caging‐group‐free photoactivatable dye with a multidrug‐binding transcriptional regulator, QacR. This system allowed transcriptional activation with mild visible light exposure, large dynamic range, and simultaneous detection of the state of the effector molecule in cell‐free and cellular conditions.

Introduction

Regulation of biological processes using photochemical compounds is a crucial method to investigate cellular activities, as it offers high spatiotemporal resolution without the need for intricate protein engineering. In particular, optical control of transcription has attracted attention because of its versatility in modulating gene expression, one of the most fundamental biological phenomena_. ^[1] Over the past decades, many efforts have been made to develop light‐mediated transcriptional regulation systems by installing photoremovable protecting groups (PPGs) on biologically active molecules such as small compounds,[ ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ ] nucleic acids,[ ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ ] and proteins.[ ²² , ²³ ] PPGs block the original function by binding to the target biomolecule and caging the site required for the biological function of the target, which is restored upon exposure to light that uncages the molecule.[ ²⁴ , ²⁵ ] Several PPGs, including o‐nitrobenzyl and coumarin derivatives, have been developed and employed for photo‐regulation of transcription. Although this approach is commonly used to achieve transcriptional regulation, there exist some limitations. PPGs often require intense and long duration of UV light exposure due to their low uncaging efficiency, which may cause undesirable damage to the cells.[ ¹ , ²⁵ ] In addition, the resulting photoproducts can be toxic. For example, commonly used o‐nitrobenzyl PPGs contain highly reactive nitroso groups. ^[25] Finally, the states of effector molecules, i.e., whether they are caged or uncaged, cannot be distinguished visually, making it impossible to monitor the activity state of the effector molecules. Hence, quantitation of transcriptionally activated cells is a significant challenge, unless the target genes encode specific proteins that produce optically detectable signals, such as green fluorescent protein, luciferase, and β‐galactosidase.[ ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁹ , ¹⁰ , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ²⁰ , ²¹ , ²² , ²³ ] Those limitations confine the application of conventional PPGs to photo‐regulation of genes in general. Therefore, it is highly desirable to develop an innovative optochemically controlled transcription module that can be activated by low‐intensity light exposure that allows simultaneous detection of the activity state of the effector molecule.

To address the technological shortcomings, here we report a new class of optical transcriptional regulation system in which a photoactivatable fluorophore functions as an effector for a transcription factor (Figure 1). Among a plethora of photoactivatable dyes, we focused on the silicon‐bridged xanthone PaX₅₆₀ recently developed by Butkevich, Hell, and co‐workers. ^[26] This photoactivatable dye is capable of switching from a non‐fluorescent Si‐xanthone, PaX₅₆₀, to a highly fluorescent Si‐pyronine, PaX₅₆₀‐Closed Form (CF), upon mild light illumination. It has been reported that PaX₅₆₀ is activated by visible light (405 nm) and exhibits high photoactivation efficiency (Φ _Act× ϵ _405nm=1410). ^[26] Furthermore, the photoconversion occurs without releasing any byproducts. Owing to its light‐induced fluorogenic nature, PaX₅₆₀ has been applied to live‐cell imaging and super‐resolution imaging.[ ²⁶ , ²⁷ ] In addition to its fluorogenicity, PaX₅₆₀ undergoes photoconversion into a cationic Si‐pyronine from a neutral Si‐xanthone. We employed the positive charge acquisition to optically regulate the activity of a multidrug‐binding transcriptional regulator, QacR. QacR found in Staphylococcus aureus is classified as a TetR‐family transcriptional regulator (TFTR). ^[28] It functions as a transcription repressor that regulates gene expression of an antibiotic efflux pump by interfering with an operator sequence, IR1.[ ²⁹ , ³⁰ ] The repression activity of QacR is mitigated upon binding of a range of cationic lipophilic dyes, such as rhodamine 6G, pyronin Y, ethidium, and crystal violet, resulting in transcriptional activation of the associated genes.[ ³¹ , ³² ] Based on these observations, we hypothesized that the use of PaX₅₆₀ as a ligand for QacR would enable photo‐regulation of the QacR activity for controlling the transcription of target genes. Moreover, we developed a P_T7/QacR promoter system by integrating this QacR–PaX₅₆₀ module into the T7 RNA polymerase (T7RNAP) expression system to demonstrate that optical control of gene expression is feasible both in a cell‐free system and live bacterial cells.

Figure 1.

Schematic of the light‐activated transcription system using the photoactivatable dye PaX₅₆₀ as an effector molecule and the IR1 operator‐binding factor QacR as a transcription regulator.

Results and Discussion

The working of our QacR–PaX₅₆₀ system relies on binding of a cationic lipophilic dye to the transcription factor QacR and positive‐charge acquisition of the dye upon photoirradiation. Structural analysis of the QacR–ligand complex revealed that binding of cationic compounds is stabilized by hydrophobic and anionic amino acid residues in the ligand binding pocket through π–π, π‐cation, and ionic interactions. ^[31] In addition, the large volume of the ligand‐binding pocket of QacR allows flexibility in the structure of ligands it can bind. Hence, we reasoned that PaX₅₆₀‐CF, the cationic form of PaX₅₆₀, could serve as a QacR inducer, albeit an artificial ligand. Before light irradiation, QacR is thought to bind to the operator sequence IR1 and inhibit RNA polymerase (RNAP) from contacting the promoter region, thereby inhibiting mRNA synthesis (Figure 1). Since QacR only recognizes cationic lipophilic compounds, transcription remained suppressed after the addition of the neutral PaX₅₆₀. However, once the illumination of blue light (405 nm) is applied, PaX₅₆₀ becomes converted to positively charged PaX₅₆₀‐CF. The charge transition leads to increased affinity for QacR, forming the QacR/PaX₅₆₀‐CF complex. This entails the detachment of QacR from the IR1 sequence, which results in the initiation of transcription by RNA polymerase.

Amongst several PaX₅₆₀ derivatives, we selected the derivative with an α‐methyl vinyl substituent (Figure 1), as it exhibits the fastest photoactivation. ^[26] PaX₅₆₀ and PaX₅₆₀‐CF were synthesized according to the published method. ^[26] As reported, an increase in the absorbance at 560 nm was observed after irradiation of 3 mW/cm², indicating the conversion into the fluorescent form (Supporting Figure S1). The absorbance measurement at 560 nm revealed that the photoconversion is completed by mild irradiation of 3 mW/cm² at 405 nm for 60 sec (Supporting Figure S2). Since cationic PaX₅₆₀‐CF shows fluorescence, the dissociation constant towards QacR was determined by fluorescent polarization assay. Titration of QacR into 50 nM of PaX₅₆₀‐CF yielded a K _d of 0.87 μM (Supporting Figure S3), which is in good agreement with the dissociation constants of reported ligands, such as Rhodamine 6G (0.20 μM), pyronine Y (5.15 μM), and ethidium bromide (0.81 μM). ^[34] This sub‐micromolar K _d value is clear evidence that PaX₅₆₀‐CF displays a strong affinity towards QacR. Further, to verify that PaX₅₆₀‐CF shows a stronger affinity towards QacR than neutral and non‐fluorescent PaX₅₆₀, we measured the dissociation constants for those compounds by isothermal titration calorimetry (ITC). Titration of QacR with increasing concentrations of PaX₅₆₀ led to no change in heat (Figure 2, left panel). On the other hand, titration with PaX₅₆₀‐CF under the identical condition yielded a dissociation constant of 3.0 μM (Figure 2, right panel). This result confirmed that cationic PaX₅₆₀‐CF has a significantly higher affinity to QacR compared to neutral PaX₅₆₀, making it an ideal light‐controllable effector for the transcription factor.

Figure 2.

Comparison of affinity for QacR between PaX₅₆₀ and PaX₅₆₀‐CF determined by ITC. Titration of QacR with the two ligands (A) PaX₅₆₀ and (B) PaX₅₆₀‐CF were implemented under identical condition, where the concentration of QacR was 30 μM and that of ligands was 225 μM. Data were analyzed using a single‐site‐binding model.

To utilize Escherichia coli BL21 (DE3) as a host strain for the light‐controlled QacR–PaX₅₆₀ transcription system, we newly developed a P_T7/QacR promoter system that allows transcriptional initiation of T7RNAP to be regulated by QacR. QacR recognizes the operator sequence IR1 located directly adjacent to a promoter sequence, thereby preventing the transition of the RNAP‐promoter complex into a transcriptionally productive state. ^[31] The sequence of IR1 and the docking site for QacR have been fully identified by DNase I footprinting assay ^[32] and QacR–DNA complex crystal structure analysis, ^[33] respectively. The crystal structure has provided us with an insight that 24‐bp sequence 5′‐TATA GACC GATC GCAC GGTC TATA‐3′ is the minimum IR1 (IR1 mini) bases required for QacR recognition. We thought to position IR1 mini downstream of the P_T7 sequence to block T7RNAP from docking to the promoter sequence. Inspired by the work by Dubendorfft and Studier on developing the P_T7/LacI promoter system, ^[33] we prepared a set of DNA constructs with varying numbers of bases between the first base transcribed in P_T7 (guanine) and the first base in IR1 mini (thymine) to explore the optimum sequence that can minimize basal transcription (Figure 3A). To determine the optimal construct, in vitro T7 transcription assay was used. Transcription reaction was carried out for 1 h at 37 °C using a thermal cycler, and RNA production was quantified by band intensities on denaturing polyacrylamide gel stained with ethidium bromide (Supporting Figure S4 and Supporting Table S3). The result indicated that transcription inhibition by QacR was enhanced with decreasing number of bases between P_T7 and IR1 mini (Figure 3B). Indeed, the RNA production level using Seq 1 as a template was reduced to 2 % of the reaction in which QacR was omitted. Shorter distance between P_T7 and IR1 mini appears to allow QacR to block T7RNAP from binding to the promoter sequence more efficiently. Previous DNase I footprinting experiments have reported that DNase could not access 2 bp each from the terminal nucleotides of IR1 mini in the presence of QacR, suggesting that QacR prevents T7RNAP from contacting the DNA sequence up to 2 bp from IR1 mini. ^[28] Seq 1, where IR1 mini is juxtaposed to P_T7, displayed the maximum transcription inhibition, and this result corresponds well with the previous report on the P_T7/LacI promoter system. ^[33] Therefore, Seq 1 was employed in our P_T7/QacR promoter system for further experiments.

Figure 3.

Optical control of in vitro T7 cell‐free transcription. A) Four constructs Seq 1–4 were examined for optimization of the number of spacer bases between the P_T7 and IR1 sequences. B) Transcriptional repression by QacR in vitro. QacR (0 or 800 nM) was added to the transcription reaction mixture containing each of the four constructs Seq 1–4 (100 nM) and T7RNAP, and the mixture was incubated at 37 °C for 1 h. The transcribed RNA was quantified by the band intensity on denaturing polyacrylamide gel stained by ethidium bromide (Supporting Figure 4 and Supporting Table 3). C) Photoactivated in vitro transcription. To the transcription reaction mixtures using the Seq 1 construct, PaX₅₆₀ (0–25 μM) was added. Photoirradiation (405 nm, 3 mW/cm² for 60 sec) was applied only to the (+) Light samples, and all samples were incubated at 37 °C for 1 h. The RNA was quantified by the same method as described above (Supporting Figure 5 and Supporting Table 4). Error bars denote standard deviation (N=3). *p<0.05, **p<0.01, ***p<0.001. Significance was evaluated using the Student t‐test.

Next, we evaluated whether the light‐mediated conversion of PaX₅₆₀ to positively charged PaX₅₆₀‐CF can inhibit the transcriptional repression activity of QacR in the in vitro T7 transcription assay. Various concentrations of PaX₅₆₀ (0, 5, 10, 15, 20, and 25 μM) were added to the transcription reaction mixture, and the samples were photoirradiated at 405 nm at 3 mW/cm² for 60 sec. The relative band intensities of the transcribed RNA on denaturing polyacrylamide gel were determined by comparing it to non‐irradiated controls. RNA production intensified in proportion to the concentration of PaX₅₆₀ for the light‐irradiated sample, while it remained relatively unchanged for the non‐irradiated controls (Figure 3C, Supporting Figure S5, and Supporting Table S4). In particular, the use of 25 μM of PaX₅₆₀ led to the highest induction (23‐fold) relative to the control. Furthermore, the levels of transcription achieved upon light activation of PaX₅₆₀ at concentrations over 20 μM approached nearly 100 % of the level observed in the absence of QacR (Figure 3C and Supporting Figure S5). The results indicated that photoactivated PaX₅₆₀‐CF induced detachment of QacR from IR1 to permit RNA synthesis by T7RNAP. Taken together, the series of experiments described above served as the proof‐of‐concept for the PaX₅₆₀‐based light‐controlled P_T7/QacR promoter system we designed.

Although QacR has been extensively characterized biochemically and structurally,[ ³⁴ , ³⁵ ] its utility in the cellular protein expression system has remained limited in part due to the infectious nature of the host bacteria, S. aureus. Thus, to investigate the operability of the system in live cells, we proceeded to install it in E.coli BL21(DE3), one of the most commonly used bacterial strains for recombinant protein production in the field of biotechnology ^[36] that carries a copy of the T7 RNAP gene on its chromosome. ^[37] To introduce the system into the host, we constructed a plasmid pUC19‐QacR‐LuxCDABE containing a luxCDABE operon ^[38] under the control of the P_T7/QacR promoter system and a QacR gene encoded in the direction opposite to the luxCDABE operon under the control of a P_lac/lacI promoter system (Figure 4A). With this system, we performed real‐time monitoring of the luciferase expression level by measuring the luminescence intensity of the bacterial culture. As anticipated, the addition of PaX₅₆₀ with light irradiation (405 nm, 3 mW/cm², 60 sec) resulted in a robust luciferase activity (Figure 4B, left panel). In contrast, the non‐photoirradiated control did not show a significant increase in the luciferase activity (Figure 4B, right panel). The luciferase activity level reached the maximum approximately 300 min after light irradiation. At 50 μM PaX₅₆₀, the induction rate was 42‐fold higher relative to the non‐illuminated control. This result verified that the P_T7/QacR system can activate gene expression in live bacterial cells upon photoactivation of PaX₅₆₀ with low‐intensity visible light illumination. The luminescence intensity was increased with the light exposure time (Supporting Figure S6), indicating the amount of luciferase expression is able to be controlled by the extent of light irradiation. With the increasing concentration of PaX₅₆₀, stagnant bacterial growth was observed among the photoirradiated samples (Supporting Figure S7). This could be due to the influence of the cationic PaX₅₆₀‐CF. In fact, when the microbial viability was examined, while the neutral PaX₅₆₀ exhibited no toxicity, PaX₅₆₀‐CF affected the growth rate at higher concentrations (Supporting Figure S8).

Figure 4.

Optical control of T7 transcription in bacterial cells. A) Schematic of the construct used for the luciferase reporter gene assay. T_rrnB: rrnB transcription terminator; qacr: QacR gene; lacO: lac operator; P_lacUV5: lacUV5 promoter; P_lacI: lacI promoter; P_T7: T7 promoter; IR1: IR1 operator; luxCDABE: luciferase operon; sfGFP: super‐folder green fluorescent protein gene; T_T7: T7 terminator. B) Light‐activated luciferase expression in E. coli BL21(DE3) with light irradiation (405 nm, 3 mW/cm², 60 sec) and C) without light irradiation. PaX₅₆₀ (0–50 μM) and 0.1 % triton X‐100 were added to the cell suspension. The cell suspension was photo‐irradiated, then incubated at 37 °C with shaking at 900 rpm for 6 h. Luminescence was recorded at the indicated time points. D) Fluorescence from the activated PaX₅₆₀ molecules (pink) provides the memory of photoirradiation. E) Fluorescence imaging of photoactivation of PaX₅₆₀ and sfGFP expression with (top row) and without (bottom row) light irradiation (405 nm, 3 mW/cm², 60 sec) in E. coli. Images were recorded 6 h after light irradiation. PaX₅₆₀ 50 μM, scale bar: 10 μm.

The distinct advantage of our system is that the activated ligand, PaX₅₆₀‐CF, shows strong fluorescence. Indeed, the fluorescence imaging data showed the photoactivated fluorescence signals of PaX₅₆₀‐CF from the bacteria after photoirradiation, indicating that the dye was internalized into the cytoplasm of the bacteria (Supporting Figure S9). We also found the fluorescence signals from PaX₅₆₀‐treated E. coli cells after washing and subsequent light irradiation (Supporting Figure S10), indicating that non‐activated PaX₅₆₀ is membrane‐permeable and can be taken up by bacterial cells. Hence, this property of PaX₅₆₀ allows simultaneous detection of the state of the effector molecule and recording of the light irradiation event (Figure 4D). To demonstrate this, we used super‐folder green fluorescent protein (sfGFP) ^[39] as a reporter protein to confirm co‐localization of PaX₅₆₀‐CF and sfGFP within the cytoplasm by confocal fluorescence imaging. Using pUC19‐QacR‐LuxCDABE as the parent plasmid, the luciferase operon was replaced with the sfGFP gene to yield pUC19‐QacR‐sfGFP. Photoirradiation on the pUC19‐QacR‐sfGFP‐transformed E. coli BL21(DE3) cell suspension led to activation of PaX₅₆₀, showing red fluorescence along with enhanced expression of sfGFP after 6 hours (Figure 4E). By contrast, cells without light illumination exhibited neither red fluorescence nor green fluorescence (Figure 4E). In the absence of photoirradiation, the photoactivatable dye remained non‐fluorescent and incapable of alleviating transcriptional suppression by QacR, resulting in no induction of sfGFP production. The clear difference in the fluorescence intensities between with and without light irradiation ensures that transcription from our P_T7/QacR system is only activated in E. coli cells that are showing red fluorescence. Thus, it is safe to conclude that the fluorescence signal from PaX₅₆₀‐CF is an important indicator that allows us to identify the activity state of the effector molecule.

Conclusion

We have developed a novel class of photochemical transcriptional regulation system using a caging‐group‐free photoactivatable dye, PaX₅₆₀, as an effector molecule and QacR as a transcriptional regulator. The key driver of this system is the light‐induced charge transition from neutral PaX₅₆₀ to positively charged PaX₅₆₀‐CF. QacR selectively recognizes the cationic form of the dye and loses its repressor function. Thus, transcription is activated upon light irradiation of PaX₅₆₀. This system is advantageous over the conventional transcriptional regulation system using caged compounds, as its activation neither requires intense UV light irradiation nor generates side products that could interfere with molecular machineries of living organisms. Further, since PaX₅₆₀‐CF is fluorescent but not PaX₅₆₀, the transcriptionally active effector molecules can be easily identified. The proof‐of‐concept for our newly developed P_T7/QacR promoter system was demonstrated through in vitro transcription assay that showed that light irradiation enhanced RNA production with a maximum of 23‐fold. Incorporation of the luciferase reporter genes into E. coli BL21(DE3) proved that the system maintained functionality in bacterial cells. Finally, fluorescence imaging showed that the fluorescence of PaX₅₆₀‐CF could serve as an indicator of the activity state of the transcription factor ligand. The system reported here created a new modality in optical transcriptional regulation that has broad applications in biomedicine, systems biology, and functional genomics. In particular, the use of the popular T7 promoter system should allow our module to be incorporated seamlessly not only into existing in vivo gene expression technologies but also into the field of cell‐free systems. Spatiotemporal activation of transcriptional/translational systems encapsulated in liposomes allows remote‐controlled production and delivery of gene products from synthetic cells. ^[21] As such, the reported system could serve as a powerful BioBrick component for developing artificial genetic circuits not only in cell‐free systems but also in live bacterial cells.

Supporting Information

The experimental procedures, the synthesis and compound characterization data, spectral data, and supportive images are shown in Supporting Information.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Nonomura T., Minoshima M., Kikuchi K., Angew. Chem. Int. Ed. 2025, 64, e202416420. 10.1002/anie.202416420

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.