Rapid detection of monkeypox virus using a CRISPR-Cas12a mediated assay: a laboratory validation and evaluation

Background The 2022 outbreak of mpox (formerly known as monkeypox) led to the spread of monkeypox virus (MPXV) in over 110 countries, demanding effective disease management and surveillance. As current diagnostics rely largely on centralised laboratory testing, our objective was to develop a simple rapid point-of-care assay to detect MPXV in clinical samples using isothermal amplification coupled with CRISPR and CRISPR-associated protein (Cas) technology. Methods In this proof-of-concept study, we developed a portable isothermal amplification CRISPR-Cas12a-based assay for the detection of MPXV. We designed a panel of 22 primer-guide RNA sets using pangenome and gene-agnostic approaches, and subsequently shortlisted the three sets producing the strongest signals for evaluation of analytical sensitivity and specificity using a fluorescence-based readout. The set displaying 100% specificity and the lowest limit of detection (LOD) was selected for further assay validation using both a fluorescence-based and lateral-flow readout. Assay specificity was confirmed using a panel of viral and bacterial pathogens. Finally, we did a blind concordance study on genomic DNA extracted from 185 clinical samples, comparing assay results with a gold-standard quantitative PCR (qPCR) assay. We identified the optimal time to detection and analysed the performance of the assay relative to qPCR using receiver operating characteristic (ROC) curves. We also assessed the compatibility with lateral-flow strips, both visually and computationally, where strips were interpreted blinded to the fluorescence results on the basis of the presence or absence of test bands. Findings With an optimal run duration of approximately 45 min from isothermal amplification to CRISPR-assay readout, the MPXV recombinase polymerase amplification CRISPR-Cas12a-based assay with the selected primer- guide set had an LOD of 1 copy per mu L and 100% specificity against tested viral pathogens. Blinded concordance testing of 185 clinical samples resulted in 100% sensitivity (95% CI 893-100) and 993% specificity (95% CI 957-100) using the fluorescence readout. For optimal time to detection by fluorescence readout, we estimated the areas under the ROC curve to be 098 at 2 min and 099 at 4 min. Lateral-flow strips had 100% sensitivity (893-100) and 986% specificity (947-100) with both visual and computational assessment. Overall, lateral-flow results were highly concordant with fluorescence-based readouts (179 of 185 tests, 968% concordant), with discrepancies associated with low viral load samples. Interpretation Our assay for the diagnosis of mpox displayed good performance characteristics compared with qPCR. Although optimisation of the assay will be required before deployment, its usability and versatility present a potential solution to MPXV detection in low-resource and remote settings, as well as a means of community-based, on-site testing.