DNA from B. fagacearum and closely related species was obtained from a previously published study (Lamarche et al., 2015) and is listed in Supplementary Table 1. To confirm the species identity of all samples, the ITS gene, recognized as the universal DNA barcode for fungi (Schoch et al., 2012), was amplified and sequenced. A genus-specific SYBR Green-based qPCR assay (Table 1) was designed in a conserved region of the ITS gene to perform an absolute quantification (Rutledge, 2011) of all the strains and to standardize their concentration when necessary.

Fresh cultures of B. fagacearum were also isolated from wood tissue of various infected Quercus spp. in the Lower Peninsula of Michigan in 2020. Isolates were grown on acidified potato dextrose agar (PDA; Difco™, Sparks, MD, USA) for 2 weeks at room temperature, and hyphal-tip purified. One mL of Lactic acid, 85% (JT Baker, Phillipsburg, NJ, USA) was added into 1L of PDA. Mycelium was then harvested, frozen in liquid nitrogen, and ground with a mortar and pestle. Subsequent DNA extraction followed the protocol described before (Parada-Rojas and Quesada-Ocampo, 2018). All samples were quantified using the SYBR Green assay and are listed in Supplementary Table 1.

To benchmark the DETECTR assay, we first developed a TaqMan^®-based real-time PCR assay. A total of 37 published ITS sequences for B. fagacearum and closely related species were aligned (Supplementary File 2). A set of primers and a probe were then designed in regions showing B. fagacearum-specific nucleotide polymorphisms. All primers were designed using previously described criteria and amplifications were conducted under the conditions described in Lamarche et al. (2015; Table 1). Positive (using B. fagacearum DNA from pure culture isolate C520, Supplementary Table 1) and negative (no DNA template) controls were included in all qPCR runs.

The DETECTR assay was performed using RPA to amplify fungal DNA with the TwistAmp Basic Kit (TwistDx, Cambridge, UK), and LbaCas12a (New England Biolabs, MA, USA) for the DETECTR part of the assay. RPA primers (Integrated DNA Technologies Inc., Coralville, IA, USA) were adapted from the TaqMan^® qPCR assay following the TwistAmp Basic kit manual and are listed in Table 1. Amplification was achieved with a few modifications to the manufacturer’s instructions. Briefly, each reaction was prepared using 29.5 μL of TwistAmp Rehydration buffer, 2.75 μL of forward and reverse primers (10 μM) and 10.8 μL of nuclease-free water. The mix was used to resuspend the reaction pellets provided with the TwistAmp Basic kit. Then, 1 μL of DNA template and 3.2 μL of 280 mM magnesium acetate were added in a final volume of 50 μL. Tubes were quickly vortexed and incubated at 37°C for 20 min, followed by inactivation at 65°C for 10 min.

The LbaCas12a (Cas12a from Lachnospiraceae bacterium ND2006) trans-cleavage assay was performed similarly to those previously described (Chen et al., 2018; Zhang et al., 2020). Cas12a targeting sequence was selected according to the presence of PAM for Lba Cas12a (TTTV) (Zetsche et al., 2015). Guide RNA and ssDNA FQ-reporter are listed in Table 1 and were designed following recommendations from NEB (New England Biolabs, MA, USA) as well as previous studies (Zetsche et al., 2015; Zhang et al., 2020). Both were ordered from Integrated DNA Technology (IDT Inc., Coralville, IA, USA). To detect the DNA target via fluorescence, Cas12a digestion reactions were prepared using 1X NEBuffer 2.1, 50 nM of LbaCas12a, 50 nM guide RNA and 250 nM FQ-reporter in a final volume of 60 μL. The mix was pre-incubated at room temperature for 10 min. Then, 6 μL of unpurified RPA product was added to the reaction and transferred to a 96-well black optical-bottom plate (Thermo Fisher Scientific, MA, USA). Fluorescence was measured at 37°C every 5 min for 1 h (excitation: 485 nm, emission: 535 nm) using the Fluoroskan Ascent plate reader (Labsystems).

To validate the specificity of the newly designed qPCR TaqMan^® for B. fagacearum, we ran the assay in triplicates on 31 B. fagacearum strains from pure cultures as well as a panel of closely related species (section “DNA from pure cultures,” Supplementary Table 1). Closely related species’ DNA samples were standardized to approximately 5,000 ITS copies with the SYBR Green assay. Since high DNA concentrations are known to sometimes inhibit amplification, we wanted to ensure that, if samples were not detected, it would be due to the specificity of the assay, and not because of potential inhibition.

In order to evaluate the sensitivity of the qPCR TaqMan^® assay, we built standard curves using serial dilutions of B. fagacearum ITS synthetic DNA target sequence, and calculated the limit of detection (LOD), which is the smallest amount of target DNA that can be detected 95% of the time with the assay (Bustin et al., 2009). Briefly, 10-fold serial dilutions of B. fagacearum ITS copies were made from a gBlocks™ gene fragment (IDT Inc., Coralville, IA, USA) comprising a segment of the B. fagacearum ITS sequence, encompassing the two TaqMan^® assay primers (Table 1) (ref GenBank: KC305152.1). The ITS concentration of each dilution was calculated using the TaqMan^® primers and SYBR Green reagents. The qPCR TaqMan^® assay was then run in triplicates for each dilution and a standard curve was obtained by plotting the Ct values against the log value of the number of ITS copies. The LOD was determined by running 20 replicates of the smallest dilution of the standard curve giving 3 positive results.

To assess the efficiency of two DNA extraction methods, standard curves were also built using DNA extracted with 2 protocols from serial dilutions of B. fagacearum conidia. A conidial suspension was prepared by growing isolate MIFCC41 on acidified PDA for 2 weeks at room temperature (Chahal et al., 2019). Conidia were harvested by flooding pure cultures with 70% ethanol, followed by two filtrations using a double layer of Miracloth (Miracloth, EMD Millipore, Billerica, MA, USA). The concentration of conidia in the suspension was determined using a hemocytometer (6.18 × 10⁷ conidia per mL). Conidia were plated onto acidified PDA to confirm their non-viability, sedimented by centrifugation, and then sent to the Tanguay Lab at the Laurentian Forestry Centre, Canada. Conidia were then resuspended in 1 mL of sterile distilled water, disaggregated by repeatedly pushing them through a 24-gauge needle, and counted with a hemocytometer. Concentration was adjusted to 10^E6 conidia per mL, and the stock solution was diluted following a series of 10-fold dilutions in water, down to 1000 conidia per mL. Aliquots of 10 μL from each concentration were dried using a SpeedVac, and then extracted using two different methods. First, to speed-up and facilitate the on-site detection process, samples were immersed in 100 μL of the QuickExtract™ Plant DNA Extraction Solution (Lucigen, WI, USA) and were immediately heated to 65°C for 6 min, followed by 98°C for 2 min. Extracted samples were stored at −20°C until use. As a comparison, we also extracted DNA from the conidia dilutions using the QIAamp DNA micro kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s recommendations. The qPCR TaqMan^® assay was run in triplicates with both sets of conidia dilutions and standard curves were obtained by plotting the Ct values against the log value of the number of conidia.

We ran the DETECTR along with the qPCR TaqMan^® (in triplicates) on the same samples to see if the source of the material would impact the performance of the assay. The list of all environmental samples, and their origins, used to validate both of our new assays, is shown in Supplementary Table 1. DNA from B. fagacearum mycelial mat samples, which ranged from development stages 0 to 5 (Chahal et al., 2021) was extracted with 200 μL of QuickExtract™ Plant DNA Extraction Solution (Lucigen, Wisconsin, USA). DNA samples from various insect vectors carrying B. fagacearum spores (Carpophilus sayi, Epuraea corticina, Glischrochilus sanguinolentus) were obtained from Lamarche et al. (2015). DNA from other insect samples not carrying B. fagacearum spores was also included to ensure the specificity of both tests (Bergeron et al., 2019).

In order to facilitate the on-site deployment of this DETECTR assay, a few modifications were made to simplify its use outside of laboratory settings (Figure 1). Most notably, each step of the protocol was carried out in a microtube of a specific color, incubation times were reduced to the minimum and reactions were lyophilized when possible. The DNA extraction step using Lucigen buffer was the same as described above, except that only 100 μL of buffer was prealiquoted in red microtubes. The RPA reaction components (TwistAmp Rehydration buffer, forward and reverse primers, MgOAc and ddH₂O) were aliquoted in green microtubes with the addition of 5% trehalose. Cas12a digestion reactions were assembled in blue microtubes with a few modifications: reactions were prepared in a final volume of 50 μl, FQ-reporter concentration was increased to 500 nM and 5% of trehalose was added. The RPA and Cas12a reactions were lyophilized overnight at −50°C in a FreeZone 2.5 Liter freeze-dryer (Labconco, Kansas City, MO, USA) and stored at −20°C until use.

Just before use, each of the RPA and Cas12a lyophilized reactions tubes were rehydrated with water (approx. 48 μL, or 3 drops) using a 1.5 mL extended fine tip transfer pipette (Samco™ Fine Tip Transfer Pipettes, 231PK, Thermo Fisher Scientific, MA, USA). A small piece of mycelium mat was inserted in red microtubes containing the QuickExtract™ buffer, and samples were then incubated at 65°C for 6 min, followed by 98°C for 2 min in a portable mini16 Thermal cycler (miniPCR Bio, Cambridge, MA, USA). Using the same transfer pipette, the extracted DNA was diluted about 10-fold when transferred to the green microtubes. The entire content of the green microtubes was then used to rehydrate the RPA reaction pellets provided with the TwistAmp Basic Kit (TwistDx, Cambridge, UK). Samples were incubated at 37°C for 10 min without further inactivation. Part of the RPA reactions (approx. 8 μL) was then transferred to the blue microtubes. Samples were incubated at 37°C for 20 min, and fluorescence was visualized with the P51™ Molecular Fluorescence Viewer (miniPCR Bio, Cambridge, MA, USA). Each sample’s fluorescence was compared to that of the negative control, which was run along the assay. The complete protocol for the on-site DETECTR assay is illustrated in Figure 1 and listed in (Supplementary File 3).

To evaluate the level of difficulty in implementing the DETECTR on-site assay with non-laboratory end users, we organized a trial consisting of 22 volunteers with various backgrounds (from non-scientific to basic/advanced laboratory skills). Each participant was given a detailed version of the on-site protocol to read beforehand (Supplementary File 3), from DNA extraction to fluorescence read-out. A demonstration as well as clarifications were offered on the day of the trial. A total of 3 randomized samples (positive fungal mats or negative fungi) as well as a negative (no template) DNA control were given to each volunteer. Participants were told to use the on-site DETECTR assay to detect the presence of B. fagacearum in their samples. Various parameters were investigated, including time to completion and accuracy of results. Suggested improvements and recommendations from participants were also considered when designing the final written version of the simplified on-site protocol (Supplementary File 3).

Analysis was conducted using R (R Core Team, 2014), RStudio (R Studio Team, 2018). The standard curves, that represent the relationships between the outcome (Ct values) and the predictors (either ITS copies or the number of extracted conidia) were evaluated by simple regression using the lm function in the R statistical package.

Performance of the DETECTR assay was compared to the qPCR reference by constructing confusion matrices for each type of samples. The python package pycm (Haghighi et al., 2018) was used to generate the tables and calculate the accuracy. The statistical comparison was performed using the mcnemar function of the statsmodels package (Seabold and Perktold, 2010) with default parameters.

We first designed a new qPCR TaqMan^® assay that could specifically detect B. fagacearum but also act as a reference for our new molecular DETECTR assay. Using closely related species listed in Supplementary Table 1 (see section “DNA from pure cultures”), our data shows amplification only for B. fagacearum DNA. There was no amplification of any sister species isolates, making this new qPCR TaqMan^® assay 100% specific in discriminating B. fagacearum among the other closely related species selected here. Of note, some very highly concentrated DNA samples from pure cultures had to be diluted first to be detectable, most probably due to the presence of inhibitors or a DNA concentration too high. Nonetheless, amplification results were proportional to the amount of DNA present in the samples.

We then assessed the sensitivity of the qPCR TaqMan^® assay and calculated its LOD. First, we ran the assay in triplicates on serial dilutions of the gBlocks™ ITS gene fragments. The standard curve obtained by plotting the Ct values against the log value of the number of ITS copies shows a square correction coefficient (R²) of 0.98 and a LOD of 2 ITS copies (Figure 2A). The LOD was determined by re-running 20 times the last dilution that gave 3 positive detection results, giving a positive detection 18/20 runs. In all cases, Ct values were proportional to the amount of template DNA used for the qPCR reactions.

The assay was also run on two series of dilutions of conidia DNA extracted with the QIAamp and Lucigen DNA extraction kits. Figure 2B shows the standard curves obtained, with R² ranging from 0.93 to 0.95. The difference between the two DNA extraction methods is clearly showed by a 3.4 Ct shift upward with the Lucigen method, making it about 10 times less sensitive than the QIAamp method, but easier to deploy.

In parallel, another goal when designing the B. fagacearum ITS qPCR TaqMan^® assay was to replace the previous qPCR detection assay that our laboratory had developed and published in 2015 (Lamarche et al., 2015), which, while being specific, lacked in sensitivity. The previous detection assay targeted the transcription elongation factor 1-a (TEF) gene instead of the ITS region. Figure 2B shows a comparison between the two detection assays using the two DNA extraction methods above mentioned. As seen clearly on the graph, results confirm that our new B. fagacearum qPCR assay allowed us to gain a downshift of 6.6 Ct (with the QIAamp extraction), which corresponds to a 97-fold increase in sensitivity when run on conidia dilutions. This result was expected since TEF is a single copy gene, while the ITS is multi-copy. The increase in sensitivity suggests that the B. fagacearum genome possesses about 100 copies of the nuclear ribosomal DNA.

The specificity of the DETECTR assay was evaluated with the same sister species isolates used with the qPCR TaqMan^® assay (see “DNA from pure cultures,” Supplementary Table 1). This new molecular test is not meant to be quantitative and therefore only determines if a sample is “positive” or “negative.” As for our qPCR TaqMan^® assay, the DETECTR test only gave a positive result for B. fagacearum, whereas every other closely related species came up negative following a fluorescence reading on a plate reader. The new DETECTR test is therefore 100% specific for B. fagacearum according to the panel of species tested.

To estimate the sensitivity of this non-quantitative test, we ran it on various B. fagacearum sample types that could be naturally found in the environment, for instance, fungal mats and insect vectors known to carry spores of the pathogen (see samples in Supplementary Table 1). We also investigated the conidia dilution set extracted with the Qiagen method, as well as DNA samples from pure cultures (Supplementary Table 1). Figure 3 represents a scatter plot graph comparing the sensitivities of the qPCR TaqMan^® and DETECTR assays for all sample types combined. Results show that the DETECTR assay does not span the same detection range as that of the reference qPCR TaqMan^® assay. As described by Bustin et al. (2009), sensitivity refers to the minimum number of copies in a sample that can be measured accurately with an assay, and it is typically expressed as the LOD. While we didn’t run the DETECTR assay on predefined amounts of copies as we did with the qPCR TaqMan^® assay to determine the LOD, plotting all the data from this study as shown in Figure 3 suggests that the DETECTR test is roughly a 1000 times less sensitive than the qPCR TaqMan^® assay. As described above, LOD is only dependant on the target copy number, which we were able to obtain for all types of samples. Sample type and DNA preparation will, however, impact the amount of target copies being present in a sample.

In order to further validate and compare our new DETECTR assay to the qPCR reference, we looked into their respective detection results of these samples. Supplementary Table 1 shows results obtained with both tests for the same sample, and Table 2 summarizes the outcome of the comparison using a confusion matrix. Results were exactly the same for both assays when investigating DNA from pure cultures and DNA from insect vectors, with positive and negative results matching perfectly, giving an accuracy of 1 in both circumstances. In a confusion table, the accuracy represents the number of all correct predictions divided by the total number of the dataset. Therefore, the best possible case is an accuracy of 1 and indeed shows here that DETECTR results for pure DNA cultures and insects thoroughly match the qPCR “reference” results. However, while the qPCR TaqMan^® assay detected a positive signal for every mycelial mat tested, the DETECTR assay missed 5 of the late-stage samples, therefore giving 5 false-negative results. This slightly lowered the accuracy of the test to 0.84375, but with a negligible difference (McNemar’s exact test, P = 0.0625). Those mats were also late-stage mats (stages 4A and 5), with the expectation that extracted DNA amount and quality would be lower than fresh mats. SYBR Green quantification of those samples indeed confirmed that very few ITS copies were present. Hence, the DETECTR test can still be considered as a reliable assay when applied to young and fresh mycelial mats.

In order to facilitate the deployment of the new DETECTR assay designed here, a few modifications were made to the original protocol, as described in section “2.6 Performance of the DETECTR and TaqMan^® qPCR assays on environmental samples.” To ensure that every end user with any experience background could handle this novel on-site detection assay, we ran a trial combining 22 participants with or without previous laboratory experience. The average time to completion was 66.6 min, where people with a scientific background fared slightly better regarding the time taken to perform the test. Accuracy of results obtained by these beta participants compared to the qPCR TaqMan^® results are shown in the bottom part of the confusion matrix (Table 2). The accuracy of the DETECTR assay was calculated at just above 80% when compared to the qPCR reference. However, a total of 12 false-negative and one false-positive results (cross contamination from a manipulation) were recorded, lowering the P-value of the McNemar test below 0.05.

The trial was also used to gather feedback from participants in order to make the on-site DETECTR test more user-friendly and easily deployable. Their suggestions and comments strongly shaped the final version of the protocol presented in (Supplementary File 3). The factors and challenges most commonly pointed out by the participants were the wording and nomenclature of the protocol (i.e., replace technical terms like “microtube, resuspend, incubation,” etc. with non-scientific terms) and the manipulation of the transfer pipette. All participants agreed that performing the protocol a second time would be much easier and that mistakes would most certainly be avoided then.