Detection and characterization of difenoconazole resistance in Stagonosporopsis citrulli from watermelon and muskmelon in Zhejiang Province of China

The watermelon and muskmelon productions are important agriculture pillar industries of Zhejiang Province in China. Difenoconazole is an imidazole-class sterol, 14-demethylase inhibitor (DMI), which has been in use for several years to control gummy stem blight (GSB) caused by Stagonosporopsis species. However, the detection and characterization of difenoconazole resistance in GSB have not been clarified. In this study, a total of 112 isolates were collected from samples of GSB on watermelon and muskmelon in five locations of Zhejiang Province, China. All of the isolates were identified as Stagonosporopsis citrulli through morphology and multiplex PCR analysis. The determination of their resistance to difenoconazole via the discriminatory dosage method showed that the total resistance frequency was 89.3%. Among the resistant sub-population, 36.6% had high-level resistance to difenoconazole (DifHR), while 46.4% and 6.3% had low- (DifLR) and moderate-level resistance (DifMR), respectively. Additionally, the difenoconazole showed a positive cross-resistance with four DMIs, i.e., tebuconazole, prochloraz, metconazole, and mefentrifluconazole, but not hexaconazole. The phenotypic analysis found that the difenoconazole resistant (DifR) isolates demonstrated attenuated ability in both the mycelial growth and sporulation compared with the difenoconazole sensitive (DifS) isolates, while there was no significant difference in pathogenicity on watermelon leaves between the DifR and DifS isolates. Further exploration of the mechanism related to difenoconazole resistance of S. citrulli isolates revealed that the resistance to difenoconazole involved four types of mutations in CYP51, i.e., G463S for DifLR, I444M, Y446H, and A464G for DifHR. No over-expression of the cyp51 gene was found in the tested DifR isolates. Furthermore, it was found that 5% of the DifR isolates were significantly more sensitive to difenoconazole after being treated with 20 μg/mL chlorpromazine hydrochloride, indicating that the efflux mechanism may be involved in these difenoconazole-resistant isolates. Together, our study results suggested that S. citrulli had a strong resistance to difenoconazole on watermelon and muskmelon, and the mutations in cyp51and changes in fungicide efflux were responsible for the emergence of difenoconazole resistance in S. citrulli.


Background
Watermelon (Citrullus lanatus) and muskmelon (Cucumis melo) are annual trailing herbs of Cucurbitaceae (Zhang et al. 2019). Globally, nearly three-quarters of the world's watermelon and muskmelon are produced in Asia, with China being the leading producer. Watermelon and muskmelon are widely cultivated in China, among which the dominant areas are the following 12 provinces: Tibet, Sichuan, Yunnan, Chongqing, Hubei, Hunan, Jiangxi, Anhui, Jiangsu, Zhejiang, Guizhou, and Shanghai. In 2019, the cultivation area and yield of watermelon and muskmelon in Zhejiang Province were 82800 hectares and 2.467 million tons, respectively, accounting for 83.81% harvest area and 88.12% yield of fruit melons in Zhejiang Province, and 4.28% harvest area and 3.26% yield in China (Zhejiang Provincial Bureau of Statistics 2019). As an important infectious disease, gummy stem blight (GSB) occurs throughout all stages of these melons' growth and development, severely impacting their production and quality. The incidence of GSB in fields is generally between 20% and 40%, while in greenhouse planting and continuous cropping, its incidence rate can increase to 80%. In GSB epidemic regions, the yield reduction can reach up to 15% (Babadoost and Zitter 2009). GSB is caused by three morphologically similar yet genetically distinct Stagonosporopsis species: S. cucurbitacearum (syn. Didymella bryoniae), S. citrulli, and S. caricae Stewart et al. 2015). A rapid PCR-based method has been developed to distinguish these three morphologically similar species. To date, the pathogen of GSB in China has not been clearly identified and analyzed.
At present, chemical control is the main approach against GSB, in which demethylation inhibitor (DMI) fungicides, such as difenoconazole, are widely employed (Shamshad et al. 2009). Due to their single action site and extensive applications, many plant pathogens have evolved various degrees of resistance (Cools et al. 2013;Price et al. 2015). Although the S. cucurbitacearum isolates from South Carolina State of the USA remain sensitive to the DMI fungicides (Keinath and Hansen 2013), low-and moderate-level resistance of GSB to difenoconazole has been reported in Chongqing and Guizhou provinces of China. However, the mechanism of difenoconazole resistance in Stagonosporopsis spp. has not been reported (Liu 2016).
DMI fungicides inhibit ergosterol biosynthesis in fungi by targeting lanosterol 14a-demethylase, which is encoded by the cyp51 gene (Price et al. 2015). The resistance mechanisms of DMI fungicides include the following: (1) mutations in the coding region of cyp51, leading to changes in the amino acid sequence of CYP51, so that it cannot be recognized by DMIs (Cools et al. 2011;Frenkel et al. 2014;Schmitz et al. 2014;Zhang et al. 2021); (2) over-expression of cyp51 (Schnabel and Jones 2001;Bolton et al. 2012;Cools et al. 2012;Ma and Tredway 2013;Zhang et al. 2021); and (3) over-expression of transporters (Kretschmer et al. 2009;Marin et al. 2016). The latter mechanism is mainly caused by one of the two efflux systems: ATP-binding cassette (ABC) transporters and major facilitator superfamily (MFS) transporters (Cools and Fraaije 2013).
Compounds regulating the activity of ABC or MFS transporters can reverse resistance in some cases due to their inhibitory effects on toxicants efflux, and these compounds have been described in medical publications as 'regulators ' , 'reversers' , 'inhibitors' , or 'chemosensitizers' to treat resistance (Robert and Jarry 2003) caused by transporters in cells. These compounds include antipsychotic drugs (such as amitriptyline, chlorpromazine, and thioridazine), immunosuppressants (such as cyclosporine A and tacrolimus), steroid or hormone analogues (such as progesterone and diethylstilbestrol), and various plant metabolites (such as curcumin, flavone, and quercetin) (Roohparvar et al. 2007). Several of these regulators have been shown to increase the toxicity of DMI against fungi, especially in the strains overexpressing ABC transporters (Schuetzer-Muehlbauer et al. 2003;Hayashi et al. 2010). However, whether the emerging DMI resistance is due to transporters-related drug efflux is a matter that has not been addressed. In this study, we isolated the GSBcausing agents in the five locations of Zhejiang Province, China, and characterized their resistance to difenoconazole. Additionally, we assessed the potential mechanisms involved in difenoconazole resistance in the S. citrulli isolates. The results of this study will provide a scientific basis for resistance management in S. citrulli to the DMI fungicides in future production.

Identification of GSB causing agents
A total of 112 single-spored isolates were isolated in 2019 from five cities in Zhejiang Province. Among them, 67 were isolated from watermelon, and 45 from muskmelon. According to the morphology of the colony and characteristics of conidia, the pathogens were first identified as Stagonosporopsis sp. Then, further identification via multiplex PCR showed that all of the isolates were S. citrulli (Additional file 1: Figure S1).

Resistance of S. citrulli to difenoconazole
The sensitivity of 112 isolates of S. citrulli to difenoconazole was determined by the method of discriminatory dosage in vitro. The results indicated that 12 isolates were sensitive to difenoconazole, which was then named Dif S . The remaining 100 isolates were resistant to difenoconazole to varying degrees, and the frequency of resistance was 89.3%. Of all the resistant isolates (Dif R ), 52 demonstrated low resistance (Dif LR ), with a frequency of 46.4%; seven isolates were moderately resistant (Dif MR ), with a frequency of 6.3%; and 41 isolates were highly resistant (Dif HR ), with a frequency of 36.6% (Table 1). In view of the two hosts, the resistance frequency of the S. citrulli isolates from watermelon was 88.1%, including 40.3% Dif LR , 4.5% Dif MR , and 43.3% Dif HR isolates; meanwhile, the respective frequencies of the Dif R , Dif LR , Dif MR , and Dif HR isolates from muskmelon were 91.1, 55.5, 8.9, and 26.7%. Taken together, these results showed that the S. citrulli isolate from both plant hosts had developed strong resistance to difenoconazole.
Growth, sporulation, and pathogenicity of Dif S and Dif R No significant differences in mycelial growth, sporulation, and pathogenicity were observed between the Dif S and Dif LR or Dif MR isolates ( Table 2). The pathogenicity of the four types of isolates on watermelon leaves also had no significant difference. However, a fitness penalty in mycelial growth and sporulation was found for the Dif HR isolates. The mycelial growth rates of Dif S and Dif HR isolates on PDA were 18.75 ± 0.25 (mean ± SD) and 17.36 ± 0.60 mm/d, respectively. There was a significant difference in mycelial growth rate among the two types of isolates (F = 5.082, P < 0.05) and the mycelial growth rates of the Dif HR isolates were significantly reduced compared with those of the Dif S isolates. The respective sporulation abilities of the Dif S and Dif HR isolates were (10.13 ± 2.89) × 10 6 and  (0.72 ± 0.62) × 10 6 conidia/mL. According to one-way ANOVA (F = 10.546, P < 0.01), it can be inferred that the sporulation ability of the Dif HR isolates was significantly lower than that of the Dif S isolates (Table 2).

cyp51 mutations in the difenoconazole-resistant isolates
The cyp51 gene in the Dif S isolates (n = 3), Dif LR isolates (n = 5), Dif MR (n = 4), and Dif HR isolates (n = 7) were amplified by PCR with primers CQ82 and CQ83, and the target DNA fragment with a length of 1516 bp was obtained and submitted to NCBI (NCBI accession No. OL677535-OL677549). Four types of nucleotide mutations, including C to G, T to C, G to A, and C to G, were found at the positions of 2551, 2555, 2603, and 2607 of the cyp51 gene coding sequence, respectively. Accordingly, the 444th amino acid was mutated from isoleucine (I) to methionine (M), the 446th from tyrosine (T) to histidine (H), the 463rd from glycine (G) to serine (S), and the 464th from alanine (A) to glycine (G). Among them, three types of mutations, namely I444M, Y446H, and A464G, were associated with the Dif HR isolates, G463S was found in the Dif LR isolates, and no mutation was observed in the Dif MR isolates (Table 3).

cyp51 expression between the Dif S and Dif R isolates
The expression levels of the cyp51 gene in Dif S (JHTGY-7 and WZTGJ-3), Dif LR (JHXGJ-3 and JHXGJ-8), Dif MR (WZXGJ-17 and NBXGY-6), and Dif HR (WZXGJ-11 and JHXGJ-35) were determined using real-time quantitative PCR. Based on the one-way ANOVA analysis, there was no over-expression of cyp51 in the resistant strains. Aside from the expression of cyp51 in Dif LR (JHXGJ-3) having decreased significantly, there was no significant difference among the tested isolates (Table 4).

Effect of chlorpromazine hydrochloride on sensitivity to difenoconazole
The intrinsic effect of chlorpromazine hydrochloride (CH) on mycelial growth was investigated, and the results showed that CH had no obvious effect on the growth of S. citrulli at a concentration of 20 μg/mL without having amended difenoconazole. When the isolates were treated with difenoconazole + CH, most isolates did not change their sensitivity phenotype to difenoconazole, indicating that Dif S cannot grow on 5 μg/mL difenoconazole or   Bold and red color mean the wild-sensitive and mutated base of the codon, respectively 5 μg/mL difenoconazole + 20 μg/mL CH, while Dif R can grow on 5 μg/mL difenoconazole or 5 μg/mL difenoconazole + 20 μg/mL CH. Only five resistant isolates (Dif R ) out of all 100 Dif R became more sensitive to difenoconazole and changed their respective sensitivity phenotypes (Table 5). Two isolates (JHXGJ-10 and NZCGY-8) changed from Dif LR isolates to Dif S , and three (HZXGJ-12, JHXGJ-24, and JHXGJ-7) changed from Dif HR isolates to Dif LR .

Discussion
Difenoconazole, a sterol demethylation inhibitor, is an important fungicide for managing watermelon and muskmelon diseases. Currently, several hundred products containing difenoconazole have been registered to control watermelon and muskmelon diseases in China. The Fungicide Resistance Action Committee (FRAC) has listed DMIs as moderate-and high-risk fungicides, and many other plant pathogens have been reported to possess varying degrees of resistance to them, including Puccinia triticina (brown rust in wheat crops) (Stammler et al. 2009) and Pyrenopeziza brassicae (light leaf spot in oilseed rape) (Helen et al. 2014). GSB is a serious disease of watermelon and muskmelon in eastern China, especially in Zhejiang Province, where difenoconazole is one of the commonly used fungicides to control the disease. It has been reported that D. bryoniae has developed resistance to difenoconazole in Chongqing and Guizhou provinces. The resistance frequency of difenoconazole reached 0.67% in Chongqing and 2.03% in Guiyang (Liu 2016). In the present study, 112 isolates of S. citrulli were sampled and isolated, and their resistance to difenoconazole was determined. The results indicated that the resistance frequency was 89.3%, significantly higher than in other regions in China and other countries (Keinath and Hansen 2013;Liu 2016). Many previous studies have shown that the adaptability of resistant isolates will decrease after most pathogens have become resistant. The cost of fungicide resistance may be reflected in the ability of the fungus to complete its life cycle and may specifically affect conidial production and dispersal, infection efficiency, mycelial growth, overwintering capabilities, and other life cycle components. It may also affect its ability to compete with fungicide-sensitive isolates in a field environment (Standish et al. 2019). Comparing the fitness of Dif S and Dif R revealed a significant decrease in both the growth rate and sporulation ability of Dif HR . The attenuated fitness has also been consistently reported in the DMI-resistant isolates of Fusarium oxysporum f. sp. fragariae and Lasiodiplodia theobromae. In F. oxysporum f. sp. fragariae, it was reported that the tebuconazole-resistant isolates had lower growth rates, sporulations, and pathogenicity than the sensitive isolates (Gu et al. 2010). Moreover, when L. theobromae from papaya in Brazil developed resistance to difenoconazole, its fitness also decreased (Li et al. 2020). Cross-resistance is defined as resistance to two or more fungicides conferred by the same resistance mechanism (Dutra et al. 2020). Generally, crossresistance is common between fungicides belonging to  the same chemical class, sharing a similar mode of action. However, this is not true in all cases (Xu et al. 2014). For the DMI fungicides, due to a range of various resistance mechanisms, such as mutations in the 14a-demethylase (cyp51) gene, overexpression of cyp51 gene, and overexpression of ABC transporters, the lack of or the variety of cross-resistance among DMIs has been reported in many phytopathogenic fungi (Xu et al. 2014). For instance, in F. graminearum, there was weak cross-resistance between difenoconazole and metconazole, yet no cross-resistance with tebuconazole, prothioconazole, and metconazole (Fu et al. 2021). Additionally, no cross-sensitivity was observed in Monilinia fructicola to prothioconazole and the other DMIs tested (Dutra et al. 2020). In F. fujikuroi, our previous reports showed that prochloraz-resistant isolates only showed cross-resistance to prothioconazole, with no cross-resistance with tebuconazole, difenoconazole, propiconazole, metconazole, hexaconazole, triadimefon, or mefentrifluconazole (Mao et al. 2020;Zhang et al. 2021;Liu et al. 2022). In this study, we found that there was no cross-resistance between difenoconazole and hexaconazole, while there was positive crossresistance between difenoconazole and other four DMIs: tebuconazole, prochloraz, metconazole, and mefentrifluconazole, indicating variability in the cross-resistance among DMIs and same DMI may even have distinct cross-resistance with other DMIs in the different fungi species. Thus, monitoring the resistance development and analyzing the specific cross patterns of pathogen populations to DMIs were important for guiding DMIs application and disease management. The most widely reported mechanism of DMI resistance in field isolates is the cyp51 mutations. These mutations lead to changes in the affinity of fungicides to this enzyme, resulting in tolerance to azoles. cyp51 mutations have been reported in some fungal plant pathogens (Cools et al. 2013;Helen et al. 2014;Wang et al. 2014). In this study, to analyze the resistance mechanisms in S. citrulli to difenoconazole, we first scan the cyp51 sequence among the Dif S , Dif LR , Dif MR , and Dif HR stains. We found four mutations within cyp51 in the isolates with different resistant levels, i.e., G463S in the Dif LR isolates, I444M, Y446H, and A464G in the Dif HR isolates. G463S and A464G had been previously detected in Stagonosporopsis species for tebuconazole resistance (Li et al. 2016), whereas I444M and Y446H were reported for the first time. Surprisingly, no mutation was observed in cyp51 of Dif MR , indicating other mechanisms except mutations in cyp51 were responsible for the emergence of Dif MR . The overexpression of the cyp51 gene is another reported mechanism involved in the reduction of sensitivity to the DMI fungicides. This mechanism has been reported in some pathogens, such as Blumeriella jaapi (cherry leaf spot) Proffer et al. 2006), Venturia inaequalis (apple scab) (Schnabel and Jones 2001), and P. brassica (Helen et al. 2014). However, the over-expression of cyp51 was not observed in the resistant isolates of S. citrulli. To investigate the involvement of efflux protein in the resistance of S. citrulli to difenoconazole, chlorpromazine hydrochloride was used. Chlorpromazine hydrochloride is a compound that can regulate the activity of ABC or MFS transporters by inhibiting the efflux of toxicants from cells (Leroux and Walker 2013). It has been previously shown that chlorpromazine can enhance the sensitivity of Botrytis cinerea to DMIs (Hayashi et al. 2010), and Leroux and Walker (2013) confirmed this finding. Although the expression level of cyp51 was not determined in all the resistant isolates, we assessed the involvement of efflux protein in difenoconazole resistance in all the S. citrulli isolates with combined treatment of difenoconazole and chlorpromazine hydrochloride. Among all the 112 isolates, 5% of Dif R isolates became more sensitive to difenoconazole after the treatment of chlorpromazine hydrochloride. Some highly resistant isolates became moderately resistant, and some lowresistant isolates became sensitive when chlorpromazine hydrochloride was added (Table 5), indicating that the efflux of fungicide was also involved in the emergence of DMIs resistance in certain S. citrulli isolates. Taken together, our results prove that the difenoconazole resistance mechanisms were complicated, and different strains, even from the same species, may have distinct mechanisms in resistance development. There are four types of mutations in the cyp51 gene as well as the changed fungicide efflux activity responsible for the emergence of difenoconazole resistance in S. citrulli.

Conclusions
In summary, GSB on watermelon and muskmelon caused by S. citrulli in Zhejiang Province, eastern China, had developed serious resistance to difenoconazole. Four types of mutations in the cyp51 gene and the changed fungicide efflux activity are responsible for the emergence of difenoconazole resistance in S. citrulli.

Origin, collection, and identification of isolates
In 2019, 16 sampling sites were selected from five main producing areas of watermelon and muskmelon in Zhejiang Province (Wenzhou, Taizhou, Ningbo, Jinhua, and Huzhou), and three to eight greenhouses were selected for each sampling site. The GSB samples were collected in each greenhouse by the random sampling method. Each sample was packed separately in a sampling bag to avoid cross-contamination, and the time and location of the sample collection were recorded. The watermelon variety was Zaojia (8424). The muskmelon varieties were netted melon, Baishami No. 1, Lübaoshi, and Cucumis Melon. The tissue separation method (Fang 1998) was used to isolate the pathogen from the samples. The samples were cut, and the infected tissue (5 × 5 mm) was incubated on PDA (200 g potato, 20 g glucose, 20 g agar, and 1 L distilled water) at 25 °C for 3 days. The mycelium at the edge of the colonies was transferred to fresh PDA to obtain a pure culture. A single-spore culture was prepared for each isolate and maintained on PDA slants at 4 °C. First, the isolates were identified by morphology (Wei 1979). Then, DNA for each isolate was extracted by the CTAB method (Guo et al. 2010) and further analyzed by multiple PCR . Information on the multiplex PCR reaction, primer sequence (CQ176-CQ181), and the PCR reaction system of Stagonosporopsis sp. is provided in Table 6 (Stewart et al. 2015). The primers (CQ176-CQ181) were synthesized by Sangon Biotech (Shanghai) Co., Ltd.

Determination of resistance to difenoconazole
Mycelium plugs with a diameter of 5 mm were obtained after 3-day-old culture plates. The mycelium plugs were placed onto PDA plates supplemented with difenoconazole at 5, 20, and 50 μg/mL (Li 2007). The control PDA plates (CK) were amended with only acetone or methanol, and the experiment was repeated three times. All plates were incubated in the dark at 25 °C for 3 days to observe the growth of the isolates. The different growth phenotypes of the measured isolates were defined as follows: The sensitive isolate (Dif S ) could not grow on PDA medium containing 5 μg/mL difenoconazole. The low-level resistant isolate (Dif LR ) grew on PDA medium with 5 μg/mL difenoconazole, yet could not grow at 20 μg/mL difenoconazole. The moderate-level resistant isolate (Dif MR ) grew at 20 μg/mL, but not at 50 μg/mL difenoconazole. The isolate with a high level of resistance (Dif HR ) was able to grow on 50 μg/mL difenoconazole (Li 2007). The resistance frequency was calculated as follows: where F R is the resistance frequency, N R is the number of resistant isolates, and N T is the total number of isolates.

Comparison of growth, conidial production, and pathogenicity between sensitive and resistant isolates
A total of 16 isolates from four sensitive types were chosen at random. The mycelial plugs (5 mm in diameter) were taken from the edge of 3-day-old PDA plates and inversely placed onto the center of the PDA media. They were placed in an incubator at 25 °C for 4 days, and the diameter of each colony was measured. The experiment was repeated three times for each isolate. The mean colony diameter (measured diameter: 5 mm) was adopted to represent mycelial growth ability (Zhang et al. 2009;Xu et al. 2014). Conidia production was determined according to the previously described sporulation method (Li 2007;Wang et al. 2017). The mycelial plugs (5 mm in diameter) were taken from the edge of a 3-day-old PDA medium and placed onto the center of potato ammonium dihydrogen phosphate agar medium (PDA 200 mL,  2 g ammonium dihydrogen phosphate). Each experiment was repeated three times for each isolate. Isolates were incubated in the dark at 25 °C for 7 days. All aerial hyphae were scraped off, and the petri dish was sealed with a sealing film and placed under black light tubes (36 W) with a 12-h light and 12-h dark illumination cycle. Another 4 days later, the conidia were washed with 20 mL ddH 2 O, and the number of conidia was counted by hemocytometer (Shi et al. 2012;Xu et al. 2014). For pathogenicity, a mycelial plug (5 mm in diameter) was taken from the edge of the 3-day-old PDA medium and placed on a 50-day-old watermelon leaf (Zaojia variety). The experiment was repeated three times with three replicate plates for each tested isolate. They were then incubated in a 25 °C incubator for 48 h. The size of the lesion (mm) was measured, and the mean size (mm) was calculated for each isolate to represent the pathogenicity (Li 2007).

Cross-resistance assay
The sensitivity assay was conducted using a mycelium growth inhibition assay. Three replicates were used for each treatment, and the tests were repeated three times. For each fungicide, the stock solutions were diluted to the desired concentrations (Additional file 2: Table S2). Colony diameters were measured after 5 days of incubation at 25 °C, and medium effective inhibitory concentration (EC 50 ) values were calculated for each isolate-fungicide combination by linear regression of the percent inhibition of mycelial growth relative to the control versus the log10 transformation for each of the fungicide concentrations. The logEC 50 of difenoconazole was used as the x-coordinate, while the logEC 50 of other DMI fungicides was individually plotted as the y-coordinate to establish the linear regression equations. Spearman rank correlation calculations were then used to assess the crossresistance between two fungicides, wherein P < 0.05 and ρ > 0.6 indicated a strong positive cross-resistance between the two fungicides (Sun et al. 2021).

Analysis of cyp51 mutations in the difenoconazole-resistant isolates
Four types of isolates were adopted at random: Dif S (n = 3), Dif LR (n = 5), Dif MR (n = 4), and Dif HR (n = 7). The complete cyp51 gene sequence (KX246903) of S. cucurbitacearum was obtained from NCBI, and the primers for the cyp51 gene of S. citrulli were designed by Primer Premier 5. The primers (CQ82 and CQ83) were synthesized by Sangon Biotech (Shanghai, China) Co., Ltd. (Table 6). The gene sequence of cyp51 was amplified by PCR using the DNA of S. citrulli as a template. PCR was conducted using a 25 μL volume reaction as follows: 12.5 μL of 2 × Taq PCR Master Mix; 1 μL of template DNA (50 ng/ μL); 1.5 μL of each pair of primers (10 μM); and 2.5 μL of ddH 2 O. The thermal cycling conditions were initial denaturation at 94 °C for 5 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1.5 min, with a final extension step at 72 °C for 10 min. PCR products of the expected size were excised from 1% agarose gel, and the amplicons were sequenced at Sangon (Shanghai, China). Finally, the DNA sequences were analyzed using DNAMAN6.0 to analyze mutations in cyp51.

RNA extraction and reverse transcription
RNA extraction was conducted according to the instructions of the UNIQ-10 column Trizol total RNA extraction kit. All materials used in the experiment were sterilized and de-enzymatic. The M-MuLV first-strand cDNA synthesis kit was used for RNA reverse transcription. A volume mixture of 12 μL was added to the test tube of the ice bath, including 0.5 μg RNA extract (100 ng/μL) and 1 μL Oligo (dT) (0.5 μg/mL), and the volume was fixed to 12 μL with RNase-free ddH 2 O. The mixture was incubated at 65 °C for 5 min, followed by an ice bath for 30 s. Reverse transcription was conducted in a volume of 20 μL, including a 12 μL reaction solution in the previous step, 4 μL 5 × reaction buffer, 2 μL dNTP mix (10 mmol/L), 1 μL M-MuLV RT (200 U/μL), and 1 μL RNase inhibitor (40 U/μL). This was mixed gently and centrifuged for 3-5 s. Then, reverse transcription was performed on the PCR instrument under the following conditions: 42 °C for 45 min, and 70 °C for 10 min. Finally, the complementary DNA was stored at −20 °C.

Real-time quantitative PCR
The SGExcel UltraSYBR Mixture kit was used for gene expression analysis in real-time quantitative PCR. Each reaction was conducted in a 20 mL volume containing 12.5 μL 2 × SGExcel Ultra SYBR mixture, 0.5 μL of each primer (0.2 μM), 1 μL cDNA, and 5.5 μL ddH 2 O. According to the instructions, the reaction process was set as follows: initial heat at 95 °C for 3 min, followed by 40 cycles of 95 °C for 15 s, and 60 °C for 20 s, then final heating at 72 °C for 25 s. The relative expression of the target gene was calculated by the 2 −ΔΔCT method. The relative expression of the cyp51 (target gene) was calculated with Actin as the reference gene and the sensitive isolate as the control group (CK). ΔC T (test sample) = C T (target gene, test sample) -C T (reference gene, test sample). ΔC T (control sample) = C T (target gene, control sample) -C T (reference gene, control sample). ΔΔC T = C T (test sample) -C T (control sample) (Kenneth and Thomas 2002).