Red pepper (Capsicum annuum L.) is one of the most economically crucial spice crops cultivated in South Korea. In 2021, approximately 92,756 tons were produced across 33,373 ha of farmland (KOSTAT, 2021). However, the production of red pepper is increasingly threatened by anthracnose, a disease primarily caused by Colletotrichum scovillei. To control this disease, several mode of action groups of fungicides have been registered. Among these, quinone outside-inhibiting (QoI) fungicides inhibit mitochondrial respiration by binding to the outer quinone oxidation site of the cytochrome bc1 complex (Becker et al., 1981). Due to their single-site mode of action, QoI fungicides are particularly susceptible to the development of resistance through specific single nucleotide polymorphisms (SNPs). Three major SNPs have been identified in plant pathogenic fungi that confer resistance to QoI fungicides: F129L (phenylalanine to leucine at position 129), G137R (glycine to arginine at position 137), and G143A (glycine to alanine at position 143) (Dorigan et al., 2023). Of these, the G143A mutation is the most commonly reported and is associated with high-level resistance, whereas F129L and G137R typically confer partial resistance. In South Korea, resistance of C. scovillei populations in red pepper fields to QoI fungicides has been linked to the G143A point mutation in the cytochrome b gene (Isa and Kim, 2022; Kim et al., 2019).
Traditionally, fungicide resistance in plant pathogens has been evaluated using an agar dilution assay, in which isolates obtained from diseased tissues are cultured on fungicide-amended media, or through molecular techniques that detect genetic mutations associated with resistance. While efficient, these methods are often impractical for large-scale resistance surveillance due to their time-consuming, labor-intensive, and costly nature, particularly when a large number of isolates must be processed. Given the importance of timely detection for effective disease management, molecular approaches have increasingly been employed to genotype pathogen populations using pooled DNA samples (Hellin et al., 2020; Zulak, 2025). Among these, quantitative sequencing (QS) has emerged as a particularly reliable and efficient approach. By enabling the direct and high-throughput detection of resistance-conferring point mutations from DNA or cDNA templates, QS offers a genotypic alternative to conventional phenotypic bioassays. Its ability to quantitatively resolve allele frequencies enables the early detection of resistance genotypes, often preceding the manifestation of phenotypic resistance (Kim et al., 2011; Kwon et al., 2008). Additionally, QS is highly reproducible, scalable, and well-suited for large-scale implementation, making it a cost-effective and practical tool for monitoring fungicide resistance at the population level across diverse geographic regions and timeframes (Cao et al., 2011; Seong et al., 2010).
In this study, we evaluated whether resistance-associated point mutations could serve as molecular markers for the rapid and accurate detection of QoI fungicide resistance in C. scovillei field populations. To this end, the applicability of QS was assessed using genomic DNA extracted from both susceptible and resistant isolates, as well as from fungal spores cultured on potato dextrose agar (PDA). Additionally, to simulate field conditions, QS was performed on DNA directly extracted from red pepper fruits that had been artificially inoculated with spore suspensions comprising varying ratios of susceptible and resistant isolates. The allele frequencies of resistance-conferring mutations determined by QS were compared with the known ratios of resistant isolate used in the inoculation. Furthermore, genomic DNA was extracted directly from symptomatic pepper fruits collected from major red pepper production regions in South Korea. The incidence of QoI fungicide resistance was determined using QS and compared with results obtained through the conventional agar dilution assay to validate the utility of QS for large-scale resistance monitoring.
Materials and Methods
Fungal isolates and fungicide
Three C. scovillei isolates (20JDS8, 20CDJ6, and JC24), originally obtained from anthracnose-infected pepper fruits, were used in this study. Isolates 20JDS8 (pyraclostrobin-sensitive) and 20CDJ6 (pyraclostrobin-resistant) were used to evaluate the applicability of QS for large-scale fungicide resistance monitoring. A single-spore isolate, C. scovillei JC24, served as a reference control to distinguish sensitive and resistant phenotypes in agar dilution assays. All isolates were cultured on PDA (DifcoTM, Becton; Dickinson and Company, Sparks, MD, USA) and maintained at 4°C. For long-term storage, cultures were overlaid with liquid paraffin and stored at −70°C. Pyraclostrobin (a.i. 20%, WG), a QoI fungicide, was used in all experiments.
Genomic DNA extraction and cyt b gene amplification
Isolates 20JDS8 and 20CDJ6 were incubated on PDA in the dark at 25°C. Mycelia were harvested, transferred into 2 mL standard microtube tubes (DeNoveTM; DeNove BioTherapeutics, Seoul, Korea), freeze-dried for 24 h, and used for DNA extraction. Total genomic DNA was extracted using a Plant SV Mini Kit (GeneAll Biotechnology, Seoul, Korea) following the manufacturer’s instructions and adjusted to 50 μg/mL using a NanoDrop spectrophotometer (K Lab Co., Ltd., Daejeon, Korea). PCR amplification of the cyt b gene was performed in 20 μL reactions containing 12 μL sterile distilled water (SDW), 4 μL EzPCR HS 5× Master Mix (Elpis Biotech, Daejeon, Korea), 1 μL of each primer (forward: Cacytb-P2, 5′-CAT AGT AAY ACA GCT TCT G-3′; reverse: Cacytb-R, 5′-GGA ATA GAT CTT AAT ATA GC-3′), and 2 μL template DNA. The PCR program consisted of initial denaturation at 94°C for 4 min; 30 cycles of 98°C for 10 s, 37°C for 30 s, and 72°C for 1 min; followed by a final extension at 72°C for 7 min. PCR products were separated on 1% agarose gels in 0.5× TBE buffer at 100 V for 40 min and visualized under UV light. Amplicons were purified using the GeneAll® Expin PCR SV Mini Kit and sequenced by Macrogen Co., Ltd. (Daejeon, Korea).
Validation of QS using genomic DNA
Genomic DNA from 20CDJ6 and 20JDS8 was standardized to 50 μg/mL and mixed at defined ratios (10:0, 9:1, 7:3, 5:5, 3:7, 1:9, and 0:10) to generate pseudo-genomic DNA pools. Each mixture was subjected to PCR, purification, sequencing, and allele frequency quantification. The mutant allele frequency (FM) was calculated as:
where hM and hW represent the peak heights of mutant and wild-type alleles, respectively (Cao et al., 2011). Peak heights were analyzed using BioEdit version 7.0 software (https://bioedit.software.informer.com/7.0/), and regression analysis was conducted against expected allele frequencies to obtain the coefficient of determination (R2).
QS using spore populations
Spores from 2-week-old PDA cultures of 20CDJ6 and 20JDS8 were harvested by gently scraping aerial mycelia. Cultures were incubated at 25°C under a 12 h light/dark cycle for 3 days to induce sporulation. Spore suspensions were filtered through double-layered gauze and adjusted to 1 × 106 spores/mL. Pseudo-spore pools were prepared by mixing resistant and sensitive spores at ratios of 10:0, 9:1, 7:3, 5:5, 3:7, 1:9, and 0:10. After centrifugation at 13,000 rpm for 30 min, pellets were subjected to DNA extraction, PCR, sequencing, and allele frequency analysis as described above.
Practical evaluation of QS using artificially inoculated pepper fruits
Surface-sterilized red pepper fruits were wound-inoculated with spore suspensions (1 × 106 spores/mL) at sensitive-to-resistant ratios of 10:0, 8:2, 5:5, 2:8, and 0:10. Three replicates per ratio were prepared, and the experiment was repeated three times. Fruits were incubated in plastic boxes at 25°C under humid conditions for 5 days, after which lids were removed and incubation continued for an additional 5 days. Disease incidence was assessed by measuring the longitudinal diameter of lesions, and statistical analyses were performed using SPSS version 26.0 (IBM Co., Armonk, NY, USA). To extract genomic DNA, infected tissues were excised, pooled, and placed in 2-mL standard microtubes (DeNoveTM) containing 0.1-mm glass beads. Tubes were sealed with Parafilm (Merck, Darmstadt. Germany), frozen at −70°C, lyophilized, and homogenized. Genomic DNA extraction, PCR, and sequencing were performed as described above. Allele frequencies were determined from sequencing chromatograms using the previously described method.
Field collection and isolation of C. scovillei
A nationwide survey was conducted in major pepper-producing regions of Korea from 2020 to 2023. Diseased red pepper fruits were collected from 83 fields in 2020, 51 fields in 2021, 72 fields in 2022, and 96 fields in 2023. Ten diseased fruits per field were sampled, and transported in cooled containers to the Laboratory of Plant Fungal Disease, Chungbuk National University. Upon arrival, fruits were incubated in a humid container for 24 h to induce sporulation, after which single-spore isolation was performed. Conidia were suspended in SDW, diluted, and plated on PDA supplemented with 300 μg/mL streptomycin. Single colonies were isolated and maintained for subsequent assays.
QS of field-collected fruits
From each field, one 5 mm2 lesion was excised from each of ten fruits, pooled, and placed in 2 mL tubes containing 0.1 mm glass beads. Samples were frozen at −70°C for 12 h, lyophilized for 24 h, and homogenized using a bullet blender (Next Advance, Troy, NY, USA). Genomic DNA extraction, PCR, and sequencing were performed as described above.
Comparison with the conventional agar dilution method
Ten isolates per field were inoculated onto PDA with or without 10.0 μg/mL of pyraclostrobin and incubated at 25°C for 5 days. The pyraclostrobin-sensitive reference isolate JC24 was included on each plate. Resistance was determined based on colony growth on pyraclostrobin-amended medium. Correlation analysis was performed between QS results and those of the agar dilution assay.
Results
QS analysis of pooled genomic DNA
Sequencing chromatograms of pooled genomic DNA derived from pyraclostrobin-sensitive and -resistant C. scovillei isolates revealed that the frequency of the G143A resistance allele increased proportionally with the proportion of resistant DNA in the mixtures. For instance, a 9:1 sensitive-to-resistant DNA mixture yielded an allele frequency of 98.3:2.7%, whereas a 1:9 mixture produced a resistant allele frequency of 100% (Fig. 1). Linear regression analysis showed an almost perfect correlation between nucleotide signal ratios and calculated allele frequencies (R2 = 0.9999; Fig. 3A).
QS analysis of spore populations
Similar results were obtained when pooled spore suspensions were used as templates. As the proportion of resistant spores increased, the frequency of the G143A allele also increased in a nearly linear manner (Fig. 2). Linear regression analysis produced a high coefficient of determination (R2 = 0.9999) and a regression equation of y = 1.0058x - 0.3454 (Fig. 3B), indicating that QS reliably reflected allele frequencies in mixed spore populations.
In vivo applicability of QS
To evaluate the applicability of QS under in vivo conditions, red pepper fruits were wound-inoculated with mixed spore suspensions at different sensitive-to-resistant ratios. Lesion lengths ranged from 20.2 ± 1.3 to 23.9 ± 1.6 mm and were not significantly different among treatments according to Duncan’s multiple range test (P > 0.05) (Figs. 4 and 5). Sequencing of DNA extracted from lesions revealed that resistance allele frequencies increased proportionally with the proportion of resistant spores in the inoculum (Fig. 6). Linear regression analysis demonstrated a perfect correlation between chromatogram signal ratios and calculated allele frequencies (R2 = 1.0000; y = 1.0001x - 0.0082), confirming the quantitative reliability of QS for in vivo samples.
Field monitoring of QoI resistance and comparison with mass agar dilution method (ADM)
Nationwide monitoring of pyraclostrobin resistance in C. scovillei populations was performed in 83, 51, 72, and 96 pepper fields in 2020, 2021, 2022, and 2023, respectively. Resistance frequencies estimated by QS were 72.0%, 75.8%, 81.6%, and 91.6% for each year, closely matching those obtained using mass ADM (77.6%, 75.3%, 79.5%, and 91.7%). Correlation analysis between QS and mass ADM revealed strong concordance, with R2 values of 0.8584, 0.9309, 0.8215, and 0.9660, respectively (Fig. 7). These results demonstrate that QS is a reliable and high-throughput alternative to conventional mass ADM for large-scale fungicide resistance monitoring.
Discussion
QS proved highly accurate and reliable tool for determining the frequency of QoI-resistant alleles in C. scovillei populations. The nearly perfect linear correlations (R2 > 0.99) between expected and measured allele frequencies from genomic DNA pools and spore pools indicate that QS accurately reflects the relative proportions of resistant and sensitive genotypes. These results are consistent with previous studies that successfully applied QS for quantifying allele frequencies in mixed pathogen populations (Bellah et al., 2023; Monsion et al., 2008).
Importantly, QS successfully quantified resistance allele frequencies in artificially inoculated pepper fruits, demonstrating its robustness under in vivo conditions. The strong correlation between chromatogram signal ratios and calculated allele frequencies suggests that host tissue does not interfere with the accuracy of the assay, consistent with previous findings that QS can be applied directly to plant tissue-derived DNA (Zhou and Mehl, 2020). This highlights the potential of QS for direct population-level monitoring without labor-intensive isolation of pure cultures. For reliable field monitoring, however, an adequate sampling size is essential to represent the population structure accurately. Unfortunately, this study did not include experiments to determine the optimal number of diseased samples to collect per field area. Based on our field experience, collecting approximately 10 infected pepper fruits per field typically provides sufficient DNA for reproducible detection and quantification of resistance allele frequencies. This sample size balances practicality and statistical confidence, although the optimal number may vary depending on disease incidence and spatial variability within the field. The development of standardized sampling guidelines will further enhance the reliability and scalability of QS-based surveillance under field conditions.
QS and QS-type approaches have already been adopted for field-level monitoring of resistance in diverse systems. For example, pyrosequencing and allele-specific qPCR have been used to quantify G143A frequencies in Cercospora sojina populations across multiple soybean fields (Zhou and Mehl, 2020), while in-field ASqPCR has been applied to detect QoI-resistant alleles in Blumeria graminis f. sp. tritici at very low frequencies (Dodhia et al., 2021). In entomology, QS has been used to monitor pyrethroid and organophosphate resistance alleles in Culex pipiens and Anopheles hyrcanus field populations (Lee et al., 2020) and to track sodium channel mutations in bed bug populations (Zhu et al., 2010). These examples demonstrate that QS is a versatile approach for resistance monitoring across pathogens and pests and reinforce its potential for high-throughput fungicide resistance surveillance in C. scovillei.
The close agreement between QS-derived resistance frequencies and those obtained using the large-scale agar dilution method (ADM) further validates QS as a reliable alternative for fungicide resistance monitoring. ADM has been widely used for resistance detection (Russell, 2002) but is time-consuming and requires culturing large numbers of isolates, which limits scalability. In contrast, QS provides rapid, quantitative allele frequency data from pooled samples, making it well-suited for high-throughput resistance surveillance.
Field monitoring from 2020 to 2023 revealed a progressive increase in the frequency of QoI-resistant C. scovillei isolates in pepper fields, reaching >90% in 2023. This trend is alarming and indicates that QoI fungicides are becoming increasingly ineffective for anthracnose management. Such findings underscore the importance of integrated disease management strategies, including the rotation of fungicides with different modes of action, adoption of cultural practices that reduce disease pressure, and deployment of resistant cultivars (Brent and Hollomon, 2007; FRAC, 2023).
Overall, this study demonstrates that QS is a robust, accurate, and efficient tool for monitoring fungicide resistance in C. scovillei populations. Whereas conventional agar dilution or large-scale agar dilution methods require at least two weeks from sample collection to the acquisition of resistance results, the QS approach described here enables completion of resistance detection within approximately 3-4 days. This 3- to 5-fold reduction in processing time provides a markedly faster diagnostic window, facilitating rapid responses to disease occurrence in agricultural fields.
Although we did not apply a correction coefficient (k) to compensate for potential allele-specific signal imbalance during Sanger sequencing, QS nonetheless exhibited strong agreement with ADM-derived resistance frequencies across all four survey years (R2 = 0.8584-0.9660). Previous QS studies have reported that Sanger-based allele frequency estimation may deviate by approximately 5-7%, or may show reduced quantitative reliability at very low (<10%) or very high (>85%) allele frequencies due to non-linear behavior of chromatogram peak ratios (Kim et al., 2011). Despite these limitations, the degree of deviation observed in Sanger-based QS is generally small and does not substantially alter the interpretation of overall resistance trends in pathogen populations. These results indicate that the uncorrected QS procedure used in this study provides sufficient accuracy, operational simplicity, and rapid turnaround time for large-scale resistance monitoring under field conditions.
Although this information is not included in the present manuscript, all previous experiments conducted in our laboratory have consistently produced results similar to those reported here, even when QS was applied to various isolates other than those used in this study. We believe this is because QoI resistance in C. scovillei arises from the same single point mutation, leading to a uniform and highly stable resistance mechanism across isolates.
Taken together, these findings highlight QS as a powerful and scalable platform for resistance surveillance and provide a practical framework for its application in routine field monitoring and integrated disease management programs targeting pepper anthracnose in agricultural production systems.
PDF Links
PubReader
ePub Link
Full text via DOI
Full text via PMC
Download Citation
Print
