LAUSR

Dear Editor, Today, the need for time-effective and cost-effective nucleic acid detection methods is still growing in fields such as human genotyping and pathogen detection. Using synthetic biomolecular components, many methods have been developed for fast nucleic acid detection; however, they may not be able to satisfy specificity, sensitivity, speed, cost and simplicity at the same time. Recently, a very promising CRISPR-based diagnostic (CRISPR-Dx) (namely SHERLOCK) was established, which was based on the collateral effect of an RNA-guided and RNAtargeting CRISPR effector, Cas13a. SHERLOCK is of high sensitivity and specificity, and is very convenient in detection of target RNA. However, to detect DNA sequences, in vitro transcription of DNA to RNA must be conducted prior to the SHERLOCK test, which could be inconvenient. In a recent study, we found that Cas12a, which belongs to the class 2 type V-A CRISPR-Cas system, performed collateral cleavage on non-targeted ssDNAs upon the formation of the Cas12a/crRNA/target DNA ternary complex. Here, with the employment of this feature, we used a quenched fluorescent ssDNA reporter (e.g., HEX-N12-BHQ1 in Supplementary Table S1) as the probe, and developed HOLMES (an one-HOur Low-cost Multipurpose highly Efficient System), which could be used for fast detection of target DNA as well as target RNA. In HOLMES, if a target DNA exists in the reaction system, the Cas12a/crRNA binary complex forms a ternary complex with the target DNA, which will then trans-cleave non-targeted ssDNA reporter in the system, illuminating the HEX fluorescence (or any other fluorescence) (Fig. 1a). We ever purified ten Cas12a proteins (Supplementary Table S3) and found all showed the ssDNA trans-cleavage activity. To find the most suitable Cas12a for HOLMES (i.e., with high signal-to-noise ratios), we tested all ten Cas12a proteins and found Lachnospiraceae bacterium ND2006 Cas12a (LbCas12a), Oribacterium sp. NK2B42 Cas12a (OsCas12a), Lachnospiraceae bacterium NC2008 Cas12a (Lb5Cas12a) and Francisella tularensis Cas12a (FnCas12a) showed good performance, among which LbCas12a was chosen for the following studies (Fig. 1b). To determine the sensitivity of HOLMES, we titrated target DNA, and found the minimum detectable concentration for Cas12a-crRNA was approximately 0.1 nM; however, when combined with PCR, the detectable concentration could be as low as 10 aM (Fig. 1c), which was comparable to the SHERLOCK system and was better than PCR alone or quantitative PCR using the SYBR Green method (Supplementary Figure S1). Therefore, to achieve higher sensitivity, PCR amplification was employed in the HOLMES test thereafter. To test whether HOLMES could discriminate singlebase differences, we made point mutations at different positions in the target DNA sequence, including both the PAM region and the guide sequences (Supplementary Figure S2a). When a full length of crRNA guide sequence (24-nt crRNA, Supplementary Table S2) was used, we found mutations in either the PAM sequences or the region of the 1st–7th bases of the guide sequence resulted in clear decline of the fluorescence signal; however, no significant difference was observed when the mutation was within the region of the 8th–18th bases (Supplementary Figure S2b), which was highly consistent with the previous report that the 5′-end seed region in the crRNA