Development of droplet digital PCR for the detection of Tilletia laevis, which causes common bunt of wheat, based on the SCAR marker derived from ISSR and real-time PCR

Common bunt of wheat caused by Tilletia laevis and/or T. caries (syn. T. tritici), is a major disease in wheat-growing regions worldwide that could lead to 80% or even total loss of production. Even though T. laevis can be distinguished from T. caries on the bases of morphology of teliospores using microscopy technique. However, molecular methods could serve as an additional method to quantify the pathogen. To develop a rapid diagnostic and quantify method, we employed the ISSR molecular marker for T. laevis in this study. The primer ISSR857 generated a polymorphic pattern displaying a 1385 bp T. laevis-specific DNA fragment. A pair of specific primers (L57F/L57R) was designed to amplify a sequence-characterized amplified region (SCAR) (763 bp) for the PCR detection assay. The primers amplified the DNA fragment in the tested isolates of T. laevis but failed in the related species, including T. caries. The detection limit of the primer set (L57F/L57R) was 5 ng/µl of DNA extracted from T. laevis teliospores. A SYBR Green I real-time PCR method for detecting T. laevis with a 100 fg/µl detection limit and droplet digital PCR with a high sensitivity (30 fg/µl detection limit) were developed; this technique showed the most sensitive detection compared to the SCAR marker and SYBR Green I real-time PCR. Additionally, this is the first study related the detection of T. laevis with the droplet digital PCR method.

sequence-characterized amplified region (SCAR) markers based on amplified fragment length polymorphism (AFLP) and inter-simple sequence repeat (ISSR) have been employed to successfully distinguish T. controversa from its related species [14][15][16][17][18] . Hence, DNA marker technology may be a powerful tool to distinguish T. laevis from other related species, especially on quantification aspects.
Compared to common PCR, real-time PCR is better with a high degree of sensitivity, specificity, repeatability, and reliability [19][20][21] , and it does not need to run gels after the reaction, save and time and eliminate the possibility of contamination. Real-time PCR is very popular in high throughput detection and quantification 22 . While, real-time PCR also has some limits 23 , such as a standard curve based on known concentration of target is necessary for getting the output data into actual values, and low accuracy of quantification will influence Cq value 24 .
However, droplet digital PCR (ddPCR), which is a sensitive technology that can amplify a highly diluted single molecule in a droplet, has the potential to improve the abovementioned limitations of real-time PCR, and the target pathogen can be detected by a fluorescent labeling probe 25,26 . Additionally, ddPCR can measure the absolute quantity of the pathogen without external nucleic acid standards. Without the need for standards, based on Poisson's distribution, positive and negative compartments are counted, and the absolute concentration of target copies in the initial sample can be determined 27 . Moreover, the final result is independent of variations in the PCR amplification efficiency, indicating that ddPCR may be more accurate, have higher repeatability, and be less prone to interlaboratory variations than real-time PCR 28 . The ddPCR method distributes the sample into thousands of independent nanoscale droplets, which removes the issues with inhibition, minimizes the deviation of reaction factors in the target samples, has the ability to accurately identify the target molecules in the presence of sufficient nontarget molecules and can calculate the accurate and original concentration of the target molecule 29 . The ddPCR method has been used for quantification, molecular identification, and evolutionary analysis; increases the amplification efficiencies and can detect the lowest concentration of the nucleic acid in the molecules 25,30 . Some reports have also mentioned that ddPCR can be more resilient to inhibitors than its non-digital counterpart 29,31,32 . Recently, ddPCR methods were successfully developed for T. controversa, a similar pathogen, with high sensitivity 33 . To date, there have been no studies using this technique for the detection of the teliospores of T. laevis.
Until now, Zhang et al. developed an AFLP-derived SCAR marker (286 bp) for T. laevis, but they only tested a limited number of similar strains and did not mention the detection limit of the SCAR marker 19 . Yao et al. developed an ISSR-derived SCAR marker (660 bp) for T. laevis with a detection limit of 0.4 ng/μl of DNA from T. laevis, and they also developed a SYBR Green I real-time PCR method based on the SCAR marker with a detection limit of 10 fg/μl of T. laevis DNA 20 . In this study, we developed a rapid and accurate method for SCAR marker detection in T. laevis, and based on the SCAR marker, we also reported that real-time PCR and droplet digital PCR with high sensitivity contribute to accurate detection. Additionally, this is the first study related to detecting the teliospore of T. laevis with the high-sensitivity ddPCR method.

Results
Specific ISSR marker screening and SCAR marker development. From 100 ISSR primers, the primer ISSR857 (5´-ACA CAC ACA CAC ACA-3´) produced a polymorphic profile (1385 bp) only in T. laevis and no polymorphic profile in any of the other investigated pathogens (Fig. 1). Based on the specific DNA sequence of T. laevis (Fig. 2), the SCAR pair of primers named L57F (5′-CGA GTG CTC TTG GTG GGA AT-3′) and L57R (5′-GCG AGG CGT TTT CAC AGT TT-3′) was designed by Primer Premier 5 for T. laevis. The primers amplified a 763 bp fragment from T. laevis. Real-time pcR. To improve the detection limit of the primers, we used real-time PCR with SYBR Green I in this study. Tenfold serial dilutions of plasmid DNA (CN = 8.29 × 10 9 -8.29 × 10 4 , 10 ng-100 fg) were used as a template (Fig. 5a). In addition, the standard curve was generated with a linear range covering 6 log units. The correlation coefficient of the standard curve reached 0.99, and the amplification efficiency was 107.3% (Fig. 5c).
To demonstrate that the amplification was specific for the SCAR marker, we performed a melting curve analysis immediately after the real-time PCR analysis. Melting curve analysis showed that the SCAR marker only had one predominant peak (Fig. 5b). These results suggested that the SYBR Green I real-time PCR detection method for T. laevis was successfully established.
Digital droplet pcR (ddpcR) detection. For ddPCR, 10,000 droplets were used, which is a precise and reliable number. More blue droplet points indicate the presence of an increased number of positive droplets in a sample and thus a greater copy number in the ddPCR product and a higher concentration of T. laevis in the DNA sample. A zero-positive droplet means there was no detection of T. laevis. The results showed that a concentrated droplet fluorescence intensity was noted in most samples with a greater number of blue droplets. Additionally, there were no blue droplets in the T. controversa samples (Fig. 6). Furthermore, the results showed that the lowest concentration of 1.5 copies/µl (30 fg/µl) was detected by ddPCR in the T. laevis DNA, and statistical analysis of the positive droplet quantities demonstrated that ddPCR was effective and successful for detection of T. laevis DNA. The analysis of the total number of droplets is shown in Fig. 7. Based on the above results, ddPCR is more sensitive and can detect the lowest concentration of DNA compared to standard PCR and real-time PCR.   19 developed a SCAR marker (286 bp) for the detection of T. laevis by AFLP, the sensitivity of the SCAR marker was not tested against T. laevis. The SCAR marker developed in this study from ISSR was 763 bp and could detect 5 ng/µl in a 25 µl PCR mixture. The 763 bp product is larger than the 286 bp product. Thus, for this procedure, it will be easier to run the gels after PCR and will save time.  www.nature.com/scientificreports/ Moreover, to further improve the sensitivity, we employed real-time PCR with SYBR Green I. Our real-time PCR results showed higher sensitivity than that of the SCAR marker with standard PCR, which was similar to other studies 33 . Recent advances in molecular detection and quantification have showed that standard PCR, SCAR markers and real-time PCR are highly efficient for pathogen detection 16. Yao et al. developed a SCAR marker for T. laevis with a detection limit of 0.4 ng/μl of DNA from T. laevis, and a SYBR Green I real-time PCR method was also successfully developed based on the SCAR marker with a detection limit of 10 fg/μl T. laevis DNA 20. In this study, the sensitivity of real-time PCR was 100 fg/µl, which was much more sensitive than that of traditional PCR detection methods (5 ng/µl).
DdPCR can achieve accurate quantification of plant pathogens without standards and is the latest and most advanced technology that is shows promise for calibration of reference materials worldwide. DdPCR can be used for the identification and quantification of pathogens, such as T. controversa 33 , which demonstrated a detection sensitivity of 2.1 copies/µl, and the results in this study showed that ddPCR could detect 30 fg/µl (1.5 copies/ µl) of T. laevis DNA. DdPCR has already been successfully used for the detection of other pathogens 31,36-38 and plant pathogens, such as Phytoplasma 39 , Erwinia amylovora and Ralstonia solanacearum 40 . Therefore, ddPCR has good potential for practical use in plant pathogen detection, especially for detection of quarantine organisms with small samples, even though running cost remains slightly above that of real-time PCR.
In summary, we developed ddPCR detection methods based on SCAR marker derived from ISSR, and realtime PCR with SYBR Green I for rapid and accurate detection of T. laevis. The obtained results from our study support the use of the ddPCR detection method in place of the SCAR marker and real-time PCR for sensitivity and accuracy of T. laevis detection. This study is the first to detect T. laevis teliospores with enhanced sensitivity of ddPCR techniques.  Table S1. DNA was extracted from urediniospores for P. striiformis f. sp. tritici, P. triticina, and P. graminis f. sp. tritici, from conidia for B. graminis, from teliospores for T. laevis, T. controversa, T. caries, U. tritici, U. hordei, U. maydis, and from vegetative hyphae for R. cerealis, F. graminearum and B. sorokiniana which were cultured on potato dextrose agar (PDA). Genomic DNA of all isolates (20 mg urediniospores, conidia or teliospores, and 3 plates of vegetative hyphae on PDA for each isolate) were extracted using a protocol reported by Liu 17 with slight   www.nature.com/scientificreports/ Cloning the species-specific DNA fragment and SCAR marker development. The specific band (1385 bp) of the T. laevis DNA generated by the primer ISSR857 (5´-ACA CAC ACA CAC ACA-3´) was excised from the gel, purified with the EasyPure Quick Gel Extraction Kit (TransGen Biotech, China), and ligated into the pMD18-T vector using a cloning kit (TaKaRa, Japan). The cloned fragment was sequenced, and a pair of SCAR marker primers (L57F:5′-CGA GTG CTC TTG GTG GGA AT-3′/L57R:5′-GCG AGG CGT TTT CAC AGT  TT-3′)

Sensitivity of the ScAR marker.
The sensitivity of the SCAR marker was tested with purified genomic DNA of T. laevis, which was serially diluted at the following concentrations: 50 ng, 25 ng, 10 ng, 5 ng, 1 ng, 100 pg 10 pg, 1 pg and 0.1 pg in 25 μl of PCR mixture. The PCR mixture, amplification procedure and agarose gel electrophoresis conditions were the same as those mentioned above.