Engineering highly thermostable Cas12b via de novo structural analyses for one-pot detection of nucleic acids

Abstract

CRISPR-Cas-based diagnostics have the potential to elevate nucleic acid detection. CRISPR-Cas systems can be combined with a pre-amplification step in a one-pot reaction to simplify the workflow and reduce carryover contamination. Here, we report an engineered Cas12b with improved thermostability that falls within the optimal temperature range (60°C-65°C) of reverse transcription-loop-mediated isothermal amplification (RT-LAMP). Using de novo structural analyses, we introduce mutations to wild-type BrCas12b to tighten its hydrophobic cores, thereby enhancing thermostability. The one-pot detection assay utilizing the engineered BrCas12b, called SPLENDID (single-pot LAMP-mediated engineered BrCas12b for nucleic acid detection of infectious diseases), exhibits robust trans-cleavage activity up to 67°C in a one-pot setting. We validate SPLENDID clinically in 80 serum samples for hepatitis C virus (HCV) and 66 saliva samples for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) with high specificity and accuracy. We obtain results in as little as 20 min, and with the extraction process, the entire assay can be performed within an hour.

Keywords: COVID-19; CRISPR; Cas12b; RT-LAMP; diagnostics; hepatitis C; one-pot detection; point-of-care diagnostics; protein engineering; thermal stability.

Graphical abstract

Figure 1

Structural prediction of BrCas12b (A) Schematics of BrCas12b trans-cleavage activity. BrCas12b forms a binary complex with its sgRNA and binds to target dsDNA, which activates its collateral ssDNA cleavage activity. (B) Predicted structures of BrCas12b derived from AlphaFold and SWISS-MODEL were spatially aligned with its most homologous Cas12b family member, BthCas12b (PDB: 5WTI). The circled regions indicate misalignment between the two predicted models when aligned to the BthCas12b structure. (C) The fitness landscape of BrCas12b was calculated by DeepDDG. Each tile represents a computationally predicted change in the free energy (ΔΔG) relative to the wild-type BrCas12b for the 20 amino acids. (D) Probability of preservation of function of three residues (F208, L795, and T874) that are likely to increase the thermostability of the BrCas12b effector, calculated by Hotspot Wizard 3.0.

Figure 2

Functional characterization of BrCas12b variants (A) Change in T_m of BrCas12b variants compared with the wild-type effector (n = 4, two technical replicates examined over two independent experiments). Error bars represent mean ± SEM. (B) Differential scanning fluorimetry plots in relative fluorescence units (RFUs) comparing the wild-type BrCas12b for selected variants screened from (A) (n = 4, two technical replicates examined over two independent experiments). (C) Differential scanning fluorimetry plots in terms of derivative of RFU in (B). Global minima of the curve were determined to be the melting point of the protein (n = 4, two technical replicates examined over two independent experiments). (D) The trans-cleavage activity of BrCas12b variants was determined at increasing temperatures. The heatmap represents the background-corrected RFU at t = 15 min (n = 2 biological replicates). (E and F) Prediction accuracy in terms of thermostability for SWISS-MODEL and AlphaFold models (n = 4, two technical replicates examined over two independent experiments).

Figure 3

Engineered BrCas12b variants exhibit exceptional stability compared with wild-type BrCas12b (A–C) Time-dependent trans-cleavage activity at different temperatures for wild-type BrCas12b, the FND (F208W/N524V/D868V) variant, and the RFND (R160E/FND) variant, respectively. The heatmap represents background-corrected mean RFU at t = 15 min (n = 2 biological replicates). (D–F) Specificity testing of wild-type BrCas12b, the FND variant, and the RFND variant against mutated activators (n = 2 biological replicates). Error bars represent mean ± SD. (G) Structural insights into the mechanism of enhanced stability of BrCas12b variants at amino acid positions 208 and 868. The top panels display the wild-type BrCas12b with F208 (left) and D868 (right), while the bottom panels depict the engineered eBrCas12b with W208 (left) and V868 (right).

Figure 4

RT-LAMP-coupled one-pot detection of BrCas12b variants (A–D) One-pot detection of BrCas12b variants at 64°C, 65°C, 66°C, and 67°C, respectively (n = 3 biological replicates). FNT, F208W/N524V/T874S; FNLD, F208W/N524V/L795I/D868V; FNPD, F208W/N524V/P763C/D868V; FNLDTA, F208W/N524V/L795I/D868V/T874S/A1015E. (E) Optimization of sucrose as an additive to the one-pot reaction at 67°C (n = 3 biological replicates). (F) One-pot detection reaction of BrCas12b variants with 200 mM sucrose at 67°C (n = 3 biological replicates). (G and H) PAM-dependent detection of the RFND variant with and without RT-LAMP reaction.

Figure 5

Clinical validation of SPLENDID in HCV-infected samples (A) Schematic of the HCV RNA genome. Single-guide RNA was designed to target the 5′ untranslated region (5′ UTR) of the virus. (B) Detailed steps of the SPLENDID assay from patient sample extraction via a magnetic-based method to the one-pot testing. (C) Fluorescence measurements of 34 HCV-1, 4 HCV-2, and 2 HCV-3 as well as 40 healthy clinical samples from the SPLENDID assay. Different genotype signal readouts were taken at t = 25 min. The reaction was run at 65°C, and this temperature was determined by screening the RT-LAMP reaction alone in the absence of the CRISPR-eBrCas12b complex. (D) Summary of clinical validation of SPLENDID in (D) in terms of sensitivity, specificity, and accuracy. (E) Receiver operating characteristic (ROC) curve of the assay at t = 25 min.