FRETting about CRISPR-Cas Assays: Dual-Channel Reporting Lowers Detection Limits and Times-to-Result

Clustered Regularly Interspaced Short Palindromic Repeats-CRISPR-Associated Protein (CRISPR-Cas) systems have evolved several mechanisms to specifically target foreign DNA. These properties have made them attractive as biosensors. The primary drawback associated with contemporary CRISPR-Cas biosensors is their weak signaling capacity, which is typically compensated for by coupling the CRISPR-Cas systems to nucleic acid amplification. An alternative strategy to improve signaling capacity is to engineer the reporter, i.e., design new signal-generating substrates for Cas proteins. Unfortunately, due to their reliance on custom synthesis, most of these engineered reporter substrates are inaccessible to many researchers. Herein, we investigate a substrate based on a fluorescein (FAM)–tetramethylrhodamine (TAMRA) Förster resonant energy-transfer (FRET) pair that functions as a seamless “drop-in” replacement for existing reporters, without the need to change any other aspect of a CRISPR-Cas12a-based assay. The reporter is readily available and employs FRET to produce two signals upon cleavage by Cas12a. The use of both signals in a ratiometric manner provides for improved assay performance and a decreased time-to-result for several CRISPR-Cas12a assays when compared to a traditional FAM–Black Hole Quencher (BHQ) quench-based reporter. We comprehensively characterize this reporter to better understand the reasons for the improved signaling capacity and benchmark it against the current standard CRISPR-Cas reporter. Finally, to showcase the real-world utility of the reporter, we employ it in a Recombinase Polymerase Amplification (RPA)–CRISPR-Cas12a DNA Endonuclease-Targeted CRISPR Trans Reporter (DETECTR) assay to detect Human papillomavirus in patient-derived samples.


Oligonucleotide details
Table S1.Sequences of the oligonucleotides used in this study.All oligonucleotides were commercially produced by Microsynth AG, Switzerland.

Characterizing Reporter Adsorption to the 384 Well Plate
To 10x HOLMES Buffer (2 µL), simulated cut reporter (FAM-CCCCCC or CCCCCC-TAMRA, or equal parts FAM-CCCCCC and CCCCCC-TAMRA) (1, 2, or 4 µL, 5 µM, UltraPure water) was added to make 250 nM, 500 nM and 1000 nM samples, respectively, and in quintuplet.All samples were then brought to a total volume of 20 µL using UltraPure water.Samples were then added to a 384 black wellplate (Corning, USA).Mineral oil (2.5 µL) was added to each well and the plate was centrifuged for 1 minute (1000 r.c.f.).The plate was then placed into a plate reader (Synergy H1, BioTek, USA) and the emission measured (Ex484/Em530 and Em583) every 2 minutes for 180 minutes.

Expression of LbCas12a Enzyme
The LbCas12a enzyme was expressed in E. coli BL21-GOLD (DE3) cells using an expression vector containing the DNA sequence for LbCas12a with an N-terminal 6xHis-tag and a C-terminal cysteine residue was (Twist Bioscience, USA).The cells were cultured in LB media at 37°C until an OD of 0.5 was reached.Next, protein expression was induced with 0.5 mM isopropyl D-thiogalactopyranoside (99%, PanReac AppliChem) and the culture was allowed to further grow for 20 hours at 20°C.The cells were harvested, resuspended in lysis buffer (50 mM Tris-HCl, 500 mM NaCl, 5% (v/v) glycerol, 1 mM TCEP, 0.5 mM PMSF, 10 mM imidazole, pH 7.5) and lysed by sonication.The recombinant protein in the soluble fraction of the lysate was isolated using immobilized metal ion affinity chromatography (Chelating Sepharose, Cytiva, USA) and then further purified by size exclusion chromatography (HiLoad 16/600 Superdex 200 pg, Cytiva, USA) using a running buffer consisting of 20 mM Tris-HCl, 250 mM NaCl, 1mM TCEP, 5% (v/v) glycerol at pH 7.5.Finally, the protein was transferred into a storage buffer prepared in nuclease-free conditions (50 mM Tris-HCl, 500 mM NaCl, 5% (v/v) glycerol, 1 mM TCEP, pH 7.5) using an Amicon-15 centrifugal filter (50 kDa MWCO, RC membrane, Merck Millipore, Germany), concentrated, aliquoted and stored at -80°C.

Clinical Sample Collection
The samples were collected by a gynaecologist using a Viba brush (Viba Brush, Rovers, Oss, The Netherlands).The cervix and the superficial vaginal canal were swabbed with the brush, which then was rinsed in Hologic ThinPrep medium (Hologic Inc., Mississauga, ON, Canada).

Clinical Sample Processing
Cervical swabs were kept in Hologic ThinPrep medium and stored at 4-8°C, then concentrated and reconstituted in 200 µl PBS with 1% IGEPAL.

Computational and Analytical Methods
All computer code along with the specific package versions and computing environment used in this analysis has been released at https://github.com/nkhosla/FRET_Ratiometric_CRISPR_Reporter.

Propagation of Error
In various analyses, we divided random variables.Each variable had an associated standard error, and thus the error was propagated to the resulting quotient.However, as the variables were often in some way experimentally or physically related, we could not assume they were independent and had zero covariance.Accordingly, a full form of the approximation of propagated uncertainty was used 5,6 i.e.

Determination of Slope
Slope was calculated by combining all data points from a replicate series into one dataset and performing a linear regression from the initial point to time t to obtain the characteristic assay slope 7 .Specifically, the linregress function from the python scipy.statspackage was used to obtain both the slope and the standard error of the calculated slope.

Time to Assay Significance Analysis
The time to assay significance, indicating the first time a statistically significant result could be read from the test, was defined as the first time, t, a given positive reached the following condition: Where P(t) and N(t) are a positive test and its corresponding negative at time t, and  % and  & are their standard deviations, respectively.

Determination of Signal:Background Ratio
Signal to background was found by dividing any given positive by its corresponding negative.That is, the negative from the same test which was also analysed the same way (single channel, ratiometric, or slope analysis).

𝑆𝐵𝑅(𝑡) = 𝑃(𝑡) 𝑁(𝑡)
It should be noted that any errors associated with the positive and negative readings were propagated as described above to calculate the SBR error.

Plotting CRISPR-Cas12a-Based Assay Data
All plots show a mean line (linearly interpolated between the data points) with shaded error ranges showing ±3 standard deviations.The shaded area is shown as a continuous region, but is also interpolated between the data points.This representation of the data was used to enhance readability, as the alternative (plotting individual data points with mean markers and error bars) proved more difficult to interpret (Figure S1).To choose a FRET pair we optimized the spectral overlap of the FRET donor and acceptor.The FAM emission spectrum was taken by exciting at 490 nm.The TAMRA excitation spectrum was measured with an emission wavelength of 583 nm.Both of these wavelengths were chosen to match the wavelengths used in the CRISPR-Cas assay using a plate reader.

Figure S1 .
Figure S1.Comparison of plotting styles.We chose to emphasize readability by showing continuous lines, while noting our actual read rate in the methods for each experiment.(A) Our adopted plotting style, showing a mean line ±3 standard deviations, both interpolated between data points.(B) The alternative, showing the same data with individual data points and error bars, along with a small mean marker.The lack of transparency makes overlapping bars hard to read, whereas adding transparency would make the thin lines hard to read.

Figure S7 .
Figure S7.Single channel readout and analysis of positive clinical samples.Single channel (Ex484/Em530) readout of all eight clinical samples (tested using a FAM-CCCCCC-BHQ reporter), showing the sample average (line) ± three standard deviations (shaded region).Each positive is plotted in comparison to the mean of all negatives (8 different clinical samples, 3 tests per sample) ± three standard deviations of the negative data.

Figure S8 .
Figure S8.Overlap of FAM Emission with TAMRA Excitation.To choose a FRET pair we optimized the spectral overlap of the FRET donor and acceptor.The FAM emission spectrum was taken by exciting at 490 nm.The TAMRA excitation spectrum was measured with an emission wavelength of 583 nm.Both of these wavelengths were chosen to match the wavelengths used in the CRISPR-Cas assay using a plate reader.
Mineral oil (2.5 µL) was added to each well and the plate centrifuged for 1 minute (1000 r.c.f.).The plate was then placed into a plate reader (BioTek Synergy H1, USA) and the emission measured (Ex484/Em530 and Em583) every 2 minutes for 180 minutes.

Table S2 .
Clinical samples as analysed by Allplex, Anyplex, and Abott commercial testing platforms.