Global food security is threatened by crop diseases caused by phytopathogens, primarily fungal pathogens. The overuse of pesticides to combat these diseases has negative environmental consequences and often fails to achieve satisfactory yield improvements. Current disease management strategies lack the precision offered by rapid, in-field molecular diagnostics. Phenotypic methods, such as symptom observation and imaging techniques, often fail to identify the specific pathogen. While nucleic acid-based molecular diagnostics offer superior pathogen identification and information about virulence and drug resistance, existing methods are complex, requiring nucleic acid amplification and laboratory settings, rendering them unsuitable for field use. This study aimed to develop a low-cost, in-field genotypic method for rapid and accurate detection of crop pathogens, bridging the gap between molecular diagnostics and conventional crop disease monitoring. This tool would enable early detection of infections, precise risk assessment, and informed fungicide selection, significantly improving crop disease control.

The paper reviews existing methods for crop disease detection. Conventional phenotypic methods like symptom observation and various imaging techniques (hyperspectral, thermographic, infrared) are mentioned, but their limitations in pathogen identification are highlighted. Molecular diagnostics based on genotypic or immunological methods are discussed, with nucleic acid-based methods showing potential for comprehensive pathogen detection. However, the complexity of existing nucleic acid-based assays, largely due to the necessity of nucleic acid amplification, restricts their applicability to laboratory settings. The need for a rapid, cost-effective, and field-deployable genotypic method that can provide information on pathogen viability and drug resistance is emphasized.

The core of the developed diagnostic tool is a nucleic acid amplification-free assay based on toehold-mediated strand displacement (TMSD) and metal ion-mediated urease catalysis. A double-stranded DNA probe (DProbe) with a specific sequence targeting pathogenic fungal RNA is designed. The presence of target RNA initiates the TMSD reaction, releasing a metal ion (Ag(I)) that inhibits urease activity. The inhibition of urease affects the hydrolysis of urea and subsequent pH change, which is visualized using a pH indicator (phenol red). The color change is observed, indicating the presence or absence of the target pathogen's RNA. The assay's principle was validated through absorbance measurements, electrophoresis, and fluorescence analysis, demonstrating its sensitivity and specificity in detecting *Puccinia striiformis* (Pst), the causal agent of wheat stripe rust. The assay's performance was further evaluated for multiplexed detection of six wheat pathogenic fungi (*Pst*, *Puccinia graminis*, *Puccinia triticina*, *Blumeria graminis*, *Fusarium graminearum*, and *Rhizoctonia cerealis*) and for detecting a plant pathogenic bacterium and a virus. Multiplex detection was achieved by integrating the individual assays onto a colorimetric paper using an origami-based paper folding strategy. A smartphone app was developed for image analysis and result interpretation. For nucleic acid extraction, a microneedle (MN) patch was employed as a rapid and reagent-free alternative to conventional methods. This patch was shown to effectively extract nucleic acids from wheat leaves within 1 minute. The entire process was designed for in-field application, requiring minimal equipment and technical expertise.

The developed smartphone-based diagnostic tool achieved high sensitivity, detecting as low as 1.0 ng/µl of Pst RNA without nucleic acid amplification. The assay successfully distinguished between six different wheat pathogenic fungi in a multiplexed format using the origami paper. The accuracy of the assay in detecting Pst infection in field-collected wheat leaf samples showed 95.3% positive predictive value and 91.0% negative predictive value compared to qPCR. The assay enabled early detection of Pst infection as early as 3 days post-inoculation, significantly earlier than symptom observation (10 days) and comparable to qPCR. The ability to differentiate between viable and dead pathogens was demonstrated by observing a correlation between the amount of viable Pst and the assay signal, while qPCR results remained similar regardless of viability. Importantly, the assay successfully identified a single-nucleotide mutation (Y134F) in the CYP51 gene, associated with triadimefon resistance in Pst isolates. This was validated by testing the isolates’ response to triadimefon, confirming the correlation between the mutation and resistance. The integration of rapid nucleic acid extraction using the MN patch, a smartphone attachment for colorimetric detection, and a user-friendly smartphone app enabled a complete sample-to-result test within approximately 10 minutes in the field. The low cost of the assay (~$0.30 per test) makes it accessible for widespread use by farmers.

The findings demonstrate a significant advancement in crop disease diagnostics. The developed platform addresses the limitations of current methods by providing a rapid, low-cost, and field-deployable solution for precise pathogen identification and resistance detection. The ability to detect viable pathogens and identify fungicide resistance is critical for optimizing disease management strategies and reducing pesticide use. Early detection of infections allows for timely interventions, minimizing yield losses. The multiplexed nature of the assay allows for comprehensive pathogen screening, improving the accuracy of diagnosis and treatment decisions. The user-friendly smartphone app simplifies the process, making it accessible to farmers with limited technical expertise. The findings are relevant for sustainable agriculture by promoting precise pesticide application and reducing environmental impact.

This study successfully developed a cost-effective, rapid, and field-deployable molecular diagnostic platform for crop diseases. The integration of toehold-mediated strand displacement, metal ion-mediated urease catalysis, colorimetric paper-based detection, and a smartphone app provides a user-friendly tool for early pathogen detection, discrimination of viable and dead fungi, identification of fungicide resistance, and guiding treatment decisions. Future research could focus on expanding the multiplexing capacity, improving quantification capabilities through advanced image processing algorithms, identifying additional drug-resistance genetic markers, and adapting the microneedle extraction method for various sample types and surfaces.

While the assay demonstrates high sensitivity and accuracy, certain limitations exist. The current colorimetric readout provides semi-quantitative data, and development of quantitative methods would enhance precision. The identification of fungicide resistance is currently limited to a specific mutation (Y134F) associated with triadimefon resistance in Pst. The expansion of the assay's capability to detect other resistance mutations is needed. The microneedle extraction method, while rapid and simple, needs further optimization for use with samples other than leaves and may not be suitable for samples with hard surfaces such as roots.