CRISPR-based diagnostics detects invasive insect pests

Abstract

Rapid identification of organisms is essential for many biological and medical disciplines, from understanding basic ecosystem processes, disease diagnosis, to the detection of invasive pests. CRISPR-based diagnostics offers a novel and rapid alternative to other identification methods and can revolutionize our ability to detect organisms with high accuracy. Here we describe a CRISPR-based diagnostic developed with the universal cytochrome-oxidase 1 gene (CO1). The CO1 gene is the most sequenced gene among Animalia, and therefore our approach can be adopted to detect nearly any animal. We tested the approach on three difficult-to-identify moth species (Keiferia lycopersicella, Phthorimaea absoluta and Scrobipalpa atriplicella) that are major invasive pests globally. We designed an assay that combines recombinase polymerase amplification (RPA) with CRISPR for signal generation. Our approach has a much higher sensitivity than real-time PCR assays and achieved 100% accuracy for identification of all three species, with a detection limit of up to 120 fM for P. absoluta and 400 fM for the other two species. Our approach does not require a sophisticated laboratory, reduces the risk of cross-contamination, and can be completed in less than 1 h. This work serves as a proof of concept that has the potential to revolutionize animal detection and monitoring.

Keywords: Keiferia lycopersicella; Phthorimaea absoluta; Scrobipalpa atriplicella; Tuta absoluta; CRISPR; Cas12a; RPA; diagnostics; genetic identification; leaf miner.

FIGURE 1

The CRISPR/Cas-based diagnostic platform for insect detection. (a) Target species included in the present study as models for developing the CRISPR/Cas assay. (b) Recombinase polymerase amplification (RPA) optimization scheme. (c) The two-pot CRISPR/Cas assay. (d) The one-pot CRISPR/Cas assay. DNA is extracted from samples with a column-based kit method (DNeasy Blood & Tissue Kit). (Fig 1b, c, d were “Created with BioRender.com”).

FIGURE 2

Screening and analytical specificity testing of Cas12a orthologs and guide RNA for the detection of insect species (a-c). Graphs depicting the fold change in fluorescence intensity with respect to the no-template control (NTC) against time for the detection of a synthetic DNA resembling the COI gene in P. absoluta (Pa), K. lycopersicella (Kl), and S. atriplicella (Sa) with Lb, As, and Er Cas12a orthologs and two different guide RNAs for each ortholog. (d) Bar graphs showing fold change with respect to the NTC for the data in panels (a-c) at time t=30 min. (e-g) Specificity testing of the CRISPR-Cas12a assay against highly similar moth species using two gRNAs designed for each species.

FIGURE 3

Limit of detection (LoD) tests for the CRISPR-Cas12a assay. (a-f) Fluorescence intensities in fold change with respect to the no-template control (NTC) in serially diluted samples containing PCR-purified COI gene fragments against the best gRNAs identified for P. absoluta (Tuta-crCOI-2), K. lycopersicella (KP-crCOI-1), and S. atriplicella (SP-crCOI-2). d-f (n=3) two out of three of samples detected in a-c (n=5) were subjected to serial dilutions around the estimated LoD for the COI gene. (g-i) Images taken under blue light of the samples detected in d-f, respectively.

FIGURE 4

Specificity of the two-pot CRISPR assay. (a-c) Fluorescence intensity in fold change with respect to the no-template control (NTC) for detecting genomic DNA. (a) Detection of P. absoluta using gRNA Tuta-crCOI-2. (b) Detection of K. lycopersicella using gRNA KP-crCOI-1. (c) Detection of S. atriplicella using gRNA SP-crCOI-2. (d-f) Representative images taken under blue light of positive samples detected in a-c, respectively, alongside corresponding NTCs. (g-i) Representation of lateral flow assay testing of the three species. Full details are in Supplementary information Fig. S10.

FIGURE 5

Testing the specificity of the one-pot CRISPR detection assay. (a-c) Fluorescence intensities in RFU in 11 different serially diluted samples containing gnomic DNA against the best gRNA’s identified for P. absoluta (Tuta-crCOI-2), K. lycopersicella (KP-crCOI-1), and S. atriplicella (SP-crCOI-2) (n=3). (d-f) Fluorescence intensity in fold change with respect to NTC for detecting genomic DNA. (d) Detection of P. absoluta. (e) Detection of K. lycopersicella. (f) Detection of S. atriplicella. The arrow in Fig. 5f indicates a false negative.