Biotin-labeled dCas9 (Bio-dCas9) preparation and validation of biotin-labeling was performed following our previous protocol (Ali et al., 2022). Briefly, the coding sequence of AviTag was cloned in-frame and downstream of the dCas9 sequence, along with the BirA gene for biotinylation of the AviTag, into the pET28a plasmid to generate pET28a-dCas9-AviTag-BirA. This plasmid was then transformed into E. coli strain BL21 (DE3) cells for protein expression. Cells were grown in 2x YT medium with 50 mg/L kanamycin sulfate at 37°C. After reaching an OD600 of 0.7 mM, 0.5 mM isopropyl β-D-thiogalactopyranoside (IPTG) for protein induction and 100 μM biotin for biotinylation of dCas9-AviTag were added. The bacterial cultures were further incubated for 16 h at 18°C. The bacteria were collected by centrifugation at 6,000 g for 20 min and then lysed in lysis buffer (2 mg/mL lysozyme, 50 mM Tris−HCl pH 8.0, 300 mM NaCl, 20 mM imidazole, 0.1% [v/v] NP-40, 1 mM PMSF, 5% [v/v] glycerol, and EDTA-free protease inhibitor cocktail tablet/50 mL [Roche, United Kingdom]). The lysate was purified by an AKTA Pure system with a HisTrap HP 5 mL affinity column (GE Healthcare) and then with a HiLoad Superdex 16/600,200 pg gel filtration column (GE Healthcare). The protein concentration was measured using a spectrophotometer (Thermo Scientific NanoDrop 8,000) and the protein was collected, concentrated, flash-frozen, and stored at −80°C.

LigRNA coding sequences were obtained as a sense strand appended with an upstream T7 promoter and an antisense strand (Supplementary Table S1). A complementary region was designed for annealing between the sense and antisense strands (ligRNA1-1 and ligRNA1-2 for ligRNA1, ligRNA2-1 and ligRNA2-2 for ligRNA2, ligRNA3-1 and ligRNA3-2 for ligRNA3, ligRNA4-1 and ligRNA4-2 for ligRNA4). The two oligonucleotides were annealed in 1X PCR buffer (−MgCl₂; Invitrogen), starting with denaturation at 95°C for 5 min, followed by 5°C ramp downsteps to 4°C. The annealed products were then amplified by PCR (with primer T7-F and individual ligRNA-R primers for ligRNA1–4) and purified (QIAquick PCR Purification Kit, QIAGEN). 1 µg of purified PCR amplicons was used as a template for in vitro transcription using Transcript Aid T7 High Yield Transcription Kit (Thermo Scientific, K0441) overnight at 37°C. The in vitro transcript was then purified with a Direct-zol RNA miniprep kit (R2050, Zymo Research). The concentration of purified ligRNAs was measured using a Nanodrop spectrophotometer (Thermo Scientific) and diluted into 6 µM working stocks. FAM-labeled amplicon for Bio-SCAN V2 execution was amplified with Target-F-FAM and Target-R primers using a TheoPCR-template as a template (Supplementary Table S2). The concentration of the 152-bp FAM-labeled amplicon was measured using a Nanodrop spectrophotometer (Thermo Scientific) and diluted into 300 ng/μL working stocks. The 370-bp target for the in vitro Cas9 cleavage experiment was amplified with Target-F-2 and Target-R primers using TheoPCR-template as a template (Supplementary Table S2). The amplified target concentration was measured using a Nanodrop spectrophotometer (Thermo Scientific) and diluted into 300 ng/μL working stocks.

For the theophylline-responsive in vitro Cas9 cleavage experiment, 300 ng of the 370-bp target amplicon, Cas9 (final concentration 50 nM), ligRNAs (final concentration 50 nM) were combined in 1X buffer 3 (20 mM HEPES pH 7.5, 150 mM KCl, 10 mM MgCl₂, 0.5 mM DTT) and incubated at 37°C for 10 min with or without 50 μM theophylline (Sigma-Aldrich, T1633-50G). After incubation, the samples were separated on 2% agarose gels and imaged using the FluorChem Q imaging System (ProteinSimple).

For ligand detection, 10 µL of theophylline (final concentration 50 µM) was incubated with 50 nM dCas9-biotin, 50 nM ligRNA1 or ligRNA2 in a 20 μL 1X buffer 1 (NEBuffer 2.1, New England Biolabs) at room temperature (RT) for 5 min to allow RNP assembly. The RNP solution (the mixture of dCas9-biotin, ligRNA1 or ligRNA2, and buffer 1 as indicated above) without theophylline was treated as the negative control. Subsequently, 1 μL FAM-labeled amplicon (300 ng) and 79 μL HybriDetect buffer (Milenia Biotec) were added to 20 μL of RNP solution to bring the final reaction volume to 100 μL. After centrifugation and vortexing, the reaction mixture was incubated at 37°C for 10 min. HybriDetect Dipsticks (Milenia Biotec) were equilibrated to room temperature (RT), placed into each tube, and removed from tubes upon the appearance of control lines. The results were obtained within 5 minutes. The appearance of both test lines and control lines represented positive samples. The appearance of only test or control lines represented invalid and negative samples, respectively.

To evaluate the effects of RNP concentration on the performance of Bio-SCAN V2, we used Bio-SCAN V2 for theophylline detection with various concentrations of the dCas9-biotin RNP. 10 µL theophylline (50 µM final concentration) was incubated with 10 nM, 30 nM, 50 nM, 75 nM, and 100 nM dCas9-biotin/ligRNA1 in 20 µL 1X buffer one solution at RT for 5 min. RNP solution without theophylline was treated as the negative control. The following steps are the same as ligand detection via Bio-SCAN V2. To test the effects of working temperature, 10 µL theophylline (50 µM final concentration) was incubated with 30 nM dCas9-biotin/ligRNA1 in 20 µL 1X buffer 1 at RT for 5 minutes. The RNP solution without theophylline was treated as the negative control. Subsequently, 1 µL FAM-labeled amplicon (300 ng) and 79 μL HybriDetect buffer was added to 20 μL RNP solution to bring the final reaction volume to 100 μL. After centrifugation and vortexing, the reaction mixture was incubated at RT, 37°C, 40°C, 45°C, or 50°C for 10 min. The following steps are the same as ligand detection via Bio-SCAN V2. To evaluate Bio-SCAN V2 function with various buffers, 10 µL theophylline (50 µM final concentration) was incubated with 30 nM dCas9-biotin/ligRNA1 in either 20 µL 1X buffer 1, 20 µL 1X buffer 2 (NEBbuffer 3.1) or 20 µL 1X buffer 3 at RT for 5 min. The RNP solution without theophylline was treated as the negative control. Bio-SCAN V2 was then executed at 45°C, following the same steps as ligand detection via Bio-SCAN V2.

To evaluate the specificity of Bio-SCAN V2, we performed Bio-SCAN V2 for the detection of caffeine (Sigma-Aldrich, C0750-5G) and xanthine (Sigma-Aldrich, X0626-5G) together with theophylline. 10 µL theophylline, caffeine, or xanthine (50 µM final concentration) was incubated with 30 nM dCas9-biotin/ligRNA1 in 20 µL 1X buffer 3 solution at RT for 5 min. RNP solution without theophylline was treated as the negative control. The following steps are the same as ligand detection via Bio-SCAN V2. To check the robustness of Bio-SCAN V2 for testing, we performed Bio-SCAN V2 for the detection of theophylline with the supplements of caffeine and Bovine serum albumin (BSA) as interferents. 10 µL theophylline (50 µM final concentration) with the supplement of 10 mM BSA and 50 µM caffeine was incubated with 30 nM dCas9-biotin/ligRNA1 in 20 µL 1X buffer 3 solution at RT for 5 min. RNP solution without theophylline was treated as the negative control. The following steps are the same as ligand detection via Bio-SCAN V2.

Theophylline (final concentrations of 0 µM, 1 µM, 2 µM, 5 µM, 10 µM, 30 µM, 100 µM, 300 µM, 1,000 µM, and 2,000 µM) was incubated with 30 nM dCas9-biotin and 30 nM ligRNA1 in 20 µL 1X buffer 3 solution at RT for 5 min. The 20 µL solutions of RNP were then incubated with 1 μL FAM-labeled amplicon (300 ng) and 79 μL HybriDetect buffer at 45°C for 5 min. HybriDetect Dipsticks were equilibrated to RT and immediately placed into each tube. The dipsticks were removed from tubes as soon as the control lines appeared. To determine theophylline concentration, linear regression between relative band intensity and the log value of theophylline concentration was calculated by GraphPad Prism 9.

The relative intensity of lateral-flow strips was analyzed by ImageJ as described by Kaminski et al. (2020). Images of lateral-flow strips were first converted to grayscale and inverted. Relative band intensity was calculated by the mean gray value of the test band divided by that of the control band.

All raw data were analyzed and visualized by GraphPad Prism 9. All numerical data were presented as mean ± standard deviation (SD). Two-tailed Student t-tests were selected for statistical analysis to detect differences between the negative control and experimental groups.

Bio-SCAN V2 was engineered from our previously reported Bio-SCAN system (Ali et al., 2022). In Bio-SCAN, we developed the biotinylated dead Cas9 (dCas9-biotin) for nucleic acid detection. During the lateral flow assay (LFA), dCas9-biotin RNP can detect a FAM-labeled amplicon, which is only produced in the presence of the nucleic acid target. For Bio-SCAN V2, we sought to design ligand-responsive sgRNA (ligRNA), which could be activated by small non-nucleic acid target molecules. The presence of target small molecules in the form of activated ligRNA is coupled with the dCas9-biotin/FAM-amplicon/αFAM antibody-AuNP complex on commercially available LFA strips for visual signal readout. To validate our CRISPR-Cas9-based small molecule detection strategy, we engineered a molecular platform called Bio-SCAN V2 for theophylline detection (Figure 1). The Bio-SCAN V2 assay requires a FAM-labeled amplicon, dCas9-biotin, ligRNA, LFA strips, and the target ligand molecule for assay actuation. FAM-labeled nucleic acids and dCas9-biotin were produced following our previous protocol (Ali et al., 2022). For ligand-induced sgRNAs, we engineered multiple ligRNAs and selected one reported sgRNA (ligRNA1) activated by theophylline, as previously shown (Kundert et al., 2019). All three ligRNAs (ligRNA2, ligRNA3, ligRNA4) were engineered by inserting the theophylline aptamer reported by Bayer et al. (Bayer and Smolke, 2005) into the hairpin structure of standard sgRNA (for Cas9) via a strand-displacement strategy. The sequences of all four ligRNAs are listed in Supplementary Table S3. The Bio-SCAN V2 assay was actuated by assembling (dCas9-biotin)-ligRNA-FAM labeled DNA in the presence and absence of target small molecules and application of the RNP-DNA complex to the αFAM antibody-AuNP embedded part of the LFA strips. The presence of the target molecule gives rise to the activation of (dCas9-biotin)-ligRNA-FAM labeled DNA complex formation and immobilization of the established (dCas9-biotin)-ligRNA-FAM labeled DNA-AuNP complex at the test line. As a result, the visual signal band is observed on the LFA strip. Unbound αFAM antibody-AuNP is detected at the control line. In the absence of target small molecules, no RNP and further (dCas9-biotin)-ligRNA-FAM labeled DNA-AuNP complex formation results in no visual band at the test line on the LFA strip.

To evaluate the ligand-induced activity of our engineered ligRNAs, we first performed an in vitro cleavage assay of a 370-bp PCR-amplified product using Cas9-ligRNAs in the presence and absence of theophylline. The PCR-amplified target was incubated with catalytically active Cas9 and four ligRNAs individually, in the presence or absence of 50 μM theophylline at 37°C for 10 min. The cleavage products were separated on a 2% agarose gel. As shown in Figure 2A, ligRNA1 and ligRNA2 showed clear Cas9-mediated DNA cleavage activity in the presence of 50 μM theophylline compared to without theophylline. Moreover, ligRNA1 demonstrated specific, high Cas9 cleavage activity in the presence of 50 μM theophylline compared to ligRNA2. The cleavage patterns of Cas9 mediated by ligRNA3 and ligRNA4 did not greatly differ in the presence or absence of 50 μM theophylline. Our results confirmed that ligRNA1 and ligRNA2 have good theophylline-responsive properties, which can potentially be developed for theophylline detection. Therefore, we selected ligRNA1 and ligRNA2 to develop the Bio-SCAN V2 system.

We next tested the Bio-SCAN V2 for theophylline detection with ligRNA1 and ligRNA2. To accomplish this, ligRNA1 or ligRNA2, dCas9-biotin, and FAM-labeled amplicon were combined and incubated with or without 50 μM theophylline at 37°C for 10 min. The reaction mixture was applied to LFA strips for visual detection without further dilution. As shown in Figure 2B, test lines were observed when executing Bio-SCAN V2 without theophylline. However, the ligRNA1 groups showed stronger test lines in the presence of 50 μM theophylline when compared to the respective negative control. On the other hand, clear bands were observed for ligRNA2 when running the Bio-SCAN V2 reaction both with and without 50 μM theophylline. Relative band intensity was quantified using ImageJ. As shown in Figure 2C, statistically significant differences were observed between ligRNA1 and ligRNA2 groups with 50 μM theophylline and without theophylline. Our results confirmed that Bio-SCAN V2 can potentially be harnessed as a tool for rapid in vitro detection of theophylline. We selected ligRNA1 for subsequent experiments, as it demonstrated better theophylline-responsive properties. However, there was still strong background observed in the ligRNA1 LFA results without theophylline. Thus, further optimization is required to adopt Bio-SCAN V2 for theophylline detection.

To enhance Bio-SCAN V2 specificity, we evaluated several parameters to achieve robust theophylline detection without background on LFA strips. First, we investigated the effect of RNP concentration on Bio-SCAN V2 performance. We carried out theophylline detection assays using different concentrations of ligRNA-(dCas9-biotin) RNP complex (10 nM, 30 nM, 50 nM, 75 nM, and 100 nM). The LFA results demonstrated that with 30 nM dCas9-biotin RNP, the Bio-SCAN V2 efficiently detected 50 μM theophylline on LFA strips, compared to no detection with 10 nM and high background in the case of 50 nM, 75 nM, 100 nM dCas9-biotin RNP (Figure 3A). The quantitative analysis clearly determined that the Bio-SCAN V2 performed best with an RNP concentration of 30 nM to detect 50 μM theophylline (Figure 3B). We next evaluated the effect of temperature on the Bio-SCAN V2 using the optimal 30 nM dCas9-biotin RNP. A Bio-SCAN V2 assay for theophylline detection was performed at room temperature, 37°C, 40°C, 45°C, and 50°C. LFA results demonstrated that Bio-SCAN V2 detected theophylline efficiently at 45°C and 50°C with lower background band appearance compared to room temperature, 37°C, and 40°C (Figures 3C, D). A working temperature of 45°C was selected for theophylline detection via Bio-SCAN V2. Next, we optimized the buffer requirements of Bio-SCAN V2. As shown in Figures 3E, F, Bio-SCAN V2 detected theophylline in buffer 3 without any background, compared to buffers 1 and 2. These results confirmed that through optimization of dCas9-biotin RNP concentration, temperature, and buffer, Bio-SCAN V2 can be efficiently adapted for the detection of small molecules including theophylline.

We selected xanthine and caffeine as the analogues to validate the specificity of Bio-SCAN V2. Among all three analytes, Bio-SCAN V2 was only able to detect theophylline. No bands were observed with samples supplemented with 50 μM xanthine or 50 μM caffeine (Figures 4A, B). Our results demonstrate that the Bio-SCAN V2 has good specificity for theophylline detection. To explore the robustness of the system, theophylline samples for Bio-SCAN V2 were supplemented with 10 mM BSA and 50 μM caffeine to simulate the co-existence of analogues in protein-abundant environments. Compared with theophylline samples without interferents, the addition of BSA and caffeine did not influence the Bio-SCAN V2-based theophylline detection, suggesting the good robustness of Bio-SCAN V2 (Figures 4C, D).

To determine the LOD of Bio-SCAN V2 for theophylline, a diagnostic assay was performed at optimized conditions to detect 0 μM, 1 μM, 2 μM, 5 μM, 10 μM, 30 μM, 100 μM, 300 μM, 1,000, and 2,000 μM theophylline. The LOD was defined as the lowest concentration showing a significant difference compared with the 0 μM theophylline groups. Our results revealed that Bio-SCAN V2 can detect as low as 2 μM theophylline (Figures 5A, B). Figure 5C illustrates the linear relationship between the log value of theophylline concentration and relative LFA band intensity. The concentration of theophylline can be calculated based on Y (relative LFA band intensity) = 0.1730*X (log value of theophylline concentration) + 0.2474.