Evaluating DFHBI-Responsive RNA Light-Up Aptamers as Fluorescent Reporters for Gene Expression

Protein-based fluorescent reporters have been widely used to characterize and localize biological processes in living cells. However, these reporters may have certain drawbacks for some applications, such as transcription-based studies or biological interactions with fast dynamics. In this context, RNA nanotechnology has emerged as a promising alternative, suggesting the use of functional RNA molecules as transcriptional fluorescent reporters. RNA-based aptamers can bind to nonfluorescent small molecules to activate their fluorescence. However, their performance as reporters of gene expression in living cells has not been fully characterized, unlike protein-based reporters. Here, we investigate the performance of three RNA light-up aptamers—F30-2xdBroccoli, tRNA-Spinach, and Tornado Broccoli—as fluorescent reporters for gene expression in Escherichia coli and compare them to a protein reporter. We examine the activation range and effect on the cell growth of RNA light-up aptamers in time-course experiments and demonstrate that these aptamers are suitable transcriptional reporters over time. Using flow cytometry, we compare the variability at the single-cell level caused by the RNA fluorescent reporters and protein-based reporters. We found that the expression of RNA light-up aptamers produced higher variability in a population than that of their protein counterpart. Finally, we compare the dynamical behavior of these RNA light-up aptamers and protein-based reporters. We observed that RNA light-up aptamers might offer faster dynamics compared to a fluorescent protein in E. coli. The implementation of these transcriptional reporters may facilitate transcription-based studies, gain further insights into transcriptional processes, and expand the implementation of RNA-based circuits in bacterial cells.

across three different media conditions (M9 Glucose, M9 Fructose, and M9 Sucrose) in both temperature settings (37°C and 30°C).The fluorescence signal obtained from each sample is normalized against the corresponding OD600 value at the maximum growth rate and presented as a relative value compared to an internal control (plasmid without expressing the aptamer) in the presence of the 160 µM DFHBI-1T.Three biological replicates were utilized, and all error bars represent the standard deviation (s.d.).Each graph contains the correlation analysis performed using GraphPrism analysis tool where the correlation coefficient, r, is calculated from the Pearson correlation coefficient and the P value indicates the significance of the correlation.

Supplementary
Fluorescence / OD600 (a.u.)  timepoints for F30-2xdBroccoli 0%-0% L-arabinose (first two columns) and 1%-0% L-arabinose (last two columns).Cells were gated for the positive (BL1-H+) and negative (BL1-H-) populations to analyse the percentage of cells activated and inactivated, respectively.Gating indicated by the horizontal bars.0%-0% sample remains off over time with ≈90% of cells within that gate.1%-0% sample lose fluorescence signal over time whereas the negative population (BL1-H-) increases over time.Supplementary Figure 11.Histograms for comparison of dynamics.Samples taken every 45 minutes for 12 timepoints for F30-2xdBroccoli 0%-1% L-arabinose (first two columns) and 1%-1% L-arabinose (last two columns).Cells were gated for the positive (BL1-H+) and negative (BL1-H-) populations to analyse the percentage of cells activated and inactivated, respectively.Gating indicated by the horizontal bars.The fluorescent signal of 0%-1% sample increases over time and the percentage of the positive population increase from 80% to 98% whereas the negative population decreases.The fluorescent signal of 1%-1% sample starts within the activated gate (BL1-H+) and increases over time.

FluorescenceSupplementary Figure 4 .
Correlation between Fluorescence Signal and Promoter Strength in RNA Aptamers Across Diverse Media and Temperature Conditions.Correlation analysis between fluorescence signals and promoter strengths for three RNA aptamers (F30-2xd Broccoli, Tornado Broccoli, and tRNA-Spinach)

Figure 5 .Supplementary Figure 6 .
Characterization of RNA light-up aptamers in microplate reader in the stationary phase (after 10h).Performance of F30-2xd Broccoli, Tornado Broccoli, and tRNA-Spinach in distinct media and temperature conditions.Each graph shows the relative fluorescence signal generated by the RNA aptamers across the promoter library under two temperature conditions (37°C in green and 30°C in yellow), both in the presence (dark green and dark yellow samples) and absence (light green and light-yellow samples) of 160 µM of fluorophore.The fluorescence signal is normalized against OD600 at the maximum growth rate and presented as a relative value compared to an internal control (plasmid without expressing the aptamer) in the presence or absence of the fluorophore.The data throughout the figure show the fluorescence produced at 10h normalised by OD600 of three biological replicates, and all error bars show the s.d.Correlation analysis results.A) Relative normalized fluorescence signals for various concentrations of L-arabinose are displayed.The fluorescence signal is normalized against OD600 at the maximum growth rate and presented as a relative value compared to an internal control (plasmid without expressing the aptamer) in the presence of the fluorophore and the corresponding inducer concentration.B) Scatterplot illustrating the relationship between the fluorescence levels of the protein reporter sfGFP and the RNA aptamers.The table shows the values for the r2 coefficient obtained from the Pearson correlation coefficient (r) and the p-value, indicating the significance of the correlation between the two studied variables.C) Relative normalized fluorescence signals for each reporter under the control of a range of constitutive promoters are presented.D) Scatterplot illustrating the relationship between the fluorescence levels of the protein reporter sfGFP and the RNA aptamers.Notably, only the F30-2xdBroccoli RNA aptamer shows a statistically significant positive correlation with the GFP fluorescence signal, with a Pearson correlation coefficient (r) of 0.9539 and a p-value = 0.0461.The sample labelled (-) represents the negative control (plasmid without expressing the aptamer).Mean and standard deviation calculations were based on results obtained from three biological replicates.

Supplementary Figure 9 .
Caption in the next page.

Supplementary Figure 9 .Supplementary Figure 10 .
Heterogeneity analysis for RNA light-up aptamers.Histograms represent three biological replicates for all RNA light-up aptamers in the presence and absence of the DFHBI-1T.Information regarding the construct, concentration of dye, replicate and cell count can be found in the tables above the histograms, with TB=Tornado Broccoli RNA aptamer, B=F30-2xdBroccoli, and S=tRNA-Spinach RNA aptamer.Histograms for comparison of dynamics.Samples taken every 45 min for 12

Fluorescence (530/30 nm) Relative Cell Count (%) Timepoint Reporter Replicate ARA initial ARA FINAL
Supplementary Figure12.Histograms for comparison of dynamics.Samples taken every 45 min for 12 timepoints for sfGFP 0%-0% L-arabinose (first two columns) and 1%-0% L-arabinose (last two columns).0%-0% sample remains off over time with no increase in the fluorescence signal.The histogram for the 1%-0% sample changes over time as the fluorescent signal starts decreasing.

Table 5 . Statistics and Tukey's multiple comparisons tests for the heterogeneity analysis of Tornado Broccoli RNA aptamer.
Geometric Mean (G.Mean), Standard Deviation (SD), Coefficient of Variation (CV) and Cell Count for three biological replicates were calculated using FlowJo Software.Statistical analysis (multiple comparisons test) was performed using Graph Prism version 9.4.1.

Table 6 . Statistics and Tukey's multiple comparisons tests for the heterogeneity analysis of F30-2xdBroccoli RNA aptamer.
Geometric Mean, Mean, Mode and Cell Count for the three biological replicates were calculated using FlowJo Software.Statistical analysis (multiple comparisons test) was performed using Graph Prism version 9.4.1.

Table 7 . Statistics and Tukey's multiple comparisons tests for the heterogeneity analysis of tRNA-Spinach RNA aptamer.
Geometric Mean, Mean, Mode and Cell Count for the three biological replicates were calculated using FlowJo Software.Statistical analysis was performed using Graph Prism version 9.4.1.

Supplementary Table 10. Normalized Cell Count for F30-2xdBroccoli RNA light-up aptamer.
Geometric Mean and standard deviation for the number of cells in the activated and inactivated populations across three biological replicates were calculated using FlowJo Software for each sample.