iSpinach: a fluorogenic RNA aptamer optimized for in vitro applications

Abstract

Using random mutagenesis and high throughput screening by microfluidic-assisted In Vitro Compartmentalization, we report the isolation of an order of magnitude times brighter mutants of the light-up RNA aptamers Spinach that are far less salt-sensitive and with a much higher thermal stability than the parent molecule. Further engineering gave iSpinach, a molecule with folding and fluorescence properties surpassing those of all currently known aptamer based on the fluorogenic co-factor 3,5-difluoro-4-hydroxybenzylidene imidazolinone (DFHBI). We illustrate the potential of iSpinach in a new sensitive and high throughput-compatible fluorogenic assay that measures co-transcriptionally the catalytic constant (kcat) of a model ribozyme.

Figure 1.

Effect of monovalent cations on the fluorescence of complexes formed between DFHBI and different DFHBI-binding RNA aptamers. 1 μM of RNA was incubated with 10 μM DFHBI in 40 mM Tris-HCl, 1 mM MgCl₂ and either 100 mM of salt (K, Na, Li or Cs) or no monovalent cation (no salt). The fluorescence of the complex was measured at 25°C. Values are the mean of three independent experiments and error bars correspond to ± 1 standard error.

Figure 2.

Microfluidic-assisted screening. (A) Experimental workflow. Steps performed on-chip (gray boxes) were distinguished from those performed off-chip (white boxes). (B) PCR droplets production. Aqueous phase supplemented with a high concentration of a orange fluorescent dye was injected into droplet generator device and 2.5 pl droplets were generated by focusing aqueous (dark orange) and oil (gray) flows. Emulsions were collected and thermocycled. (C) Droplets fusion. Small PCR droplets were reinjected into droplet-fusion device and spaced by a stream of oil. 16 pl droplets containing In Vitro Transcription (IVT, light orange) mixture supplemented with DFHBI were concomitantly produced and synchronized with PCR droplets. Pairs of droplets were then fused when passing in between a ground-connected electrode (gnd, in black) and an electrode to which tension (pos, in red) was applied. (D) Droplets sorting. After incubation, emulsions were reinjected into a Fluorescence Activated Droplet Sorting device and the fluorescence of each droplet read at a detection point (blue arrow and blue line on the micrograph). Based on the fluorescence signal, droplets of interest (green) were deflected into sort channel by applying tension to one of the electrode (pos, in red) whereas non-fluorescent droplets (orange) flowed into the waste channel. (E) Typical fluorescence profile of screened emulsion. The analysis of DFHBI green fluorescence and Texas-Red orange fluorescence allowed identifying the different populations composing the emulsion. Indeed, using orange fluorescence signal, IVT droplets fused to single PCR droplets (populations 2–5) were easily discriminated from unfused (population 1) and double fused (population 6) IVT droplets. Green fluorescence resulting from EvaGreen intercalation allowed discriminating droplets containing amplified DNA (population 4) from droplet resulting from fusion with an initially empty PCR droplet (population 2). Finally, the stronger DFHBI green fluorescence allowed discriminating droplets containing non-fluorogenic (population 2) and highly fluorogenic aptamers (population 5, red dashed boxed). Population 5 was gated and corresponding droplets sorted. (F) Evolution profiles of SpiSel-derived mutants. Gene libraries obtained after mutagenesis (in red) or screening steps (in gray) were transcribed, the RNAs purified and their fluorogenic properties assayed in the presence of the salt used for the selection. Fluorescence values were normalized to that of SpiSel (black circle) in the same conditions. Screenings performed in potassium (triangles) were distinguished from those performed in sodium (squares). Values are the mean of two independent experiments and error bars correspond to ± 1 standard error.

Figure 3.

Engineering of improved aptamers. (A–D) Secondary structure models of aptamers. Molecules are represented according to the model introduced in(21). Mutations are shown in red and numbering adapted to molecule length. (E) Effect of J1/2 deletion and A42U mutation on the fluorescence of aptamer 4–68. (F) Effect of min68/49 miniaturization on complex fluorescence in the presence or absence of constant regions. Mutant min68/49 was shortened by 3 (−3) or 6 (−6) base pairs. The fluorescence of constructs was tested in the presence and absence of constant regions. (G) Effect of double mutations on min68/49–3 fluorescence. The nature of the monovalent cation used is indicated by the color code shown in the inset. Assays were performed with 1 μM RNA and 10 μM DFHBI at 37°C. Values are the mean of three independent experiments and error bars correspond to ± 1 standard error.

Figure 4.

Comparison of Spinach2 and iSpinach. (A) Secondary structure models. RNAs were represented according to the model introduced in (21). Mutations are shown in red and numbering adapted to molecule length. (B) Aptamers properties. Main parameters including Folding Efficiency (F.E.), melting temperature of the complex and dissociation constant (K_d) of each RNA as well as the Relative Molar Fluorescence of DFHBI were determined at 25°C. Values are the mean of three independent experiments and error bars correspond to ± 1 standard error. Excitation/emission spectra are shown on Supplementary Figure S4.

Figure 5.

Real-time monitoring of RNA synthesis and ribozyme activity. (A) Constructs used. DNA fragments coding for iSpinach, Spinach2 or iXm1 ribozyme were combined in different ways and placed under the control of T7 RNA polymerase promoter sequence (Pro_T7). (B) Real-time ribozyme activity monitoring. The different constructs were in vitro transcribed in the presence of ribozyme fluorogenic substrate (S21-Atto) and the fluorescence monitored at 37°C. Product generation rate and uncatalyzed reaction rate (k_uncat) were determined as the slope of the linear phase of reaction respectively in the presence and absence of ribozyme. (C) Real-time transcription monitoring. The in vitro transcription mixture used in B. was supplemented with DFHBI (instead of S21-Atto) and the fluorescence monitored at 37°C. The synthesis rate (σ) was determined as the slope of the linear phase.