Monitoring and analysis of rice pathogen Ustilaginoidea virens isolates with resistance to sterol demethylation inhibitors in China

Rice false smut (RFS), caused by Ustilaginoidea virens (Cooke) Takah, is an important fungal disease of rice. In China, sterol demethylation inhibitors (DMIs) are common fungicides used to control RFS. In a previous study, we detected two propiconazole-resistant U. virens isolates in 2015 in Huai’an city, Jiangsu Province, China. In the current study, we detected six propiconazole-resistant isolates out of 180 U. virens isolates collected from rice fields in Jiangsu Province in 2017, and found they were from three different places (Xuzhou, Huai’an and Jintan). All these six propiconazole-resistant isolates were cross-resistant to three other sterol demethylation inhibitor (DMI) fungicides, i.e. difenoconazole, tebuconazole, and epoxiconazole. Among them, two isolates (2017–61 and 2017–170) had high fitness. Through sequencing and RT-qPCR analysis, we found that the expression levels of CYP51 and its encoded protein were significantly increased in the propiconazole-resistant isolates with a “CC” insertion mutation upstream of the CYP51 coding region compared to the propiconazole-sensitive isolates. In addition, propiconazole stimulated CYP51 expression in all isolates. Propiconazole also stimulated the accumulation of CYP51 protein in propiconazole-sensitive isolates and propiconazole-resistant isolates without mutation, but not in propiconazole-resistant isolates with the “CC” mutation. According to JASPAR database analysis, the predicated functional binding sites for propiconazole-resistant isolates with a “CC” insertion mutation and propiconazole-sensitive isolates were different. Given the high fitness of the propiconazole-resistant isolates, the development of resistance to DMIs in U. virens should be monitored. Furthermore, we speculated that the over-expression of CYP51 may contribute to DMI resistance in U. virens with the “CC” insertion mutation.

are particularly important for the management of DMI resistance in RFS.
DMIs are broad-spectrum fungicides that inhibit ergosterol biosynthesis by targeting sterol 14α-demethylase CYP51, a key enzyme in sterol biosynthesis. Homologous genes of CYP51 have been found in phytopathogenic fungi Fusarium graminearum, Magnaporthe oryzae, and Aspergillus spp. (Mellado et al. 2001;Yin et al. 2009;Yan et al. 2011). However, only one CYP51 was identified in the U. virens genome . The common mechanisms of fungal resistance to DMI include mutations in the target CYP51 gene; overexpression of the CYP51 gene, and enhanced efflux pumps that are encoded by ATPbinding cassette (ABC) transporter (Ma et al. 2006;Kretschmer et al. 2009;Cools et al. 2010Cools et al. , 2012Wang et al. 2015;Tsao et al. 2016). In F. graminearum, resistance to DMIs was not only related to the mutations of CYP51A, but also related to the ABC transporter (Ammar et al. 2013;Duan et al. 2018). In M. fructicola, a transposable element 'Mona' was inserted in the 113 bp upstream of MfCYP51, which determined the resistance to DMI fungicides Chen et al. 2017). In U. virens, up-regulation of CYP51 and increased ergosterol biosynthesis were reported to be associated with propiconazole resistance (Zhou et al. 2019). However, due to the lack of resistant field isolates, the resistance mechanism of U. virens to DMI fungicides requires further investigation.
Thus, the major objectives of the study are to (i) monitor the resistance of U. virens to propiconazole in Jiangsu Province of China, (ii) assess the resistance risk of U. virens to DMI fungicides, and (iii) investigate the possible mechanisms of DMI fungicide resistance in U. virens.
Sensitivity test of the preliminarily screened resistant isolates of U. virens to propiconazole As shown in Fig. 1, the propiconazole-sensitive isolates 2017-11 and 2017-48 could not grow on PSA plates containing 0.4 μg/mL of propiconazole. However, the six aforementioned propiconazole-resistant isolates could grow on PSA plates with 10 μg/mL of propiconazole. The propiconazole sensitivity (EC 90 ) of these resistant isolates significantly increased compared to that of the sensitive isolate 2017-11. All these resistant isolates were lowly resistant to propiconazole with resistance factor (RF) values less than 10.0 (Table 2).

Cross-resistance analysis of propiconazole-resistant U. virens isolates to other DMIs
Based on the EC 50 values, the isolates 2017-11 and 2017-48 that were sensitive to propiconazole were also sensitive to epoxiconazole, difenoconazole, and tebuconazole (Table 3). In contrast, all six propiconazoleresistant isolates were resistant to these three tested DMIs with resistance factors (RF) ranging from 2.2 to 4.6, 2.0 to 6.0, and 1.7 to 3.7, respectively. It was worth noting that the RF values of 2017-6 to difenoconazole and tebuconazole, and 2017-170 to tebuconazole were less than 1, but the EC 90 values of them were much higher than that of the sensitive isolates (Table 3).
Mycelial growth, sporulation and pathogenicity of propiconazole-resistant U. virens isolates The mycelial growth of all six propiconazole-resistant isolates was slower than that of the propiconazolesensitive isolate 2017-11. However, there was no significant difference in mycelial growth between resistant isolates and the sensitive isolate 2017-48, and the mycelial  (Table 4). Two resistant isolates, 2017-61 and 2017-170, exhibited increased spore production in potato sucrose broth (PSB) in comparison to the propiconazole-sensitive isolate 2017-11. In addition, the virulence of the field-resistant isolate 2017-61 was stronger than that of the sensitive isolates 2017-11 and 2017-48 (Table 4).
Sequence and expression analysis of the CYP51 and its encoded protein in propiconazole-resistant U. virens isolates A multisequence alignment showed no mutations in the coding region of CYP51, a reported target gene of DMIs, for all six propiconazole-resistant isolates (data not shown). However, an insertion of two bases (CC) was found 154-bp upstream of the CYP51 coding region in 2017-61, 2017-170 and 2017-179 (Fig. 2a). The Gen-Bank accession numbers were MT701730, MT701731, and MT701732 for 2017-61, 2017-170, and 2017-179, respectively. The expression levels of CYP51 and its encoded protein were significantly increased in these three resistant isolates compared to the sensitive isolate 2017-11 (Fig. 2b, c). In addition, the expression of CYP51 was significantly up-regulated by propiconazole in all isolates. However, the protein accumulation of CYP51 was increased in sensitive isolate and also in resistant isolates without the 'CC' mutation treated with 1 μg/mL of propiconazole, but not in propiconazoleresistant isolates with the "CC" mutation ( Fig. 2c).

Prediction of functional binding sites in the promoter of CYP51
The promoter region predication of CYP51 in U. virens was made using the online BDGP database. In the promoter region, 118 and 132 putative transcription factor binding sites were predicted in sequence A (without the 'CC' insertion mutation) and B (with the 'CC' insertion mutation) using JASPAR online software, respectively (Additional file 1: Tables S1 and S2). There were 20 different putative binding sites between sequences A and B. Among them, 16 binding sites were specific to sequence B with 'CC' insertion mutations, and the binding abilities of two sites of sequence B were greater than that of sequence A (Table 5).

Discussion
DMI fungicides such as Armure (30% difenoconazolepropiconazole) and Horizon (tebuconazole) have been  widely used for controlling RFS in the past two decades in China. Wang et al. (2015) obtained DMI-resistant mutants of U. virens by UV-irradiation in the laboratory. However, two propiconazole-resistant U. virens isolates were detected in Huai'an, Jiangsu Province, in 2015 (Zhou et al. 2019). In the current study, six out of 180 U. virens isolates were found to be resistant to propiconazole. In addition to Huai'an, resistant isolates were also detected in Xuzhou and Jintan in 2017. The frequency of resistant isolates in Huai'an, Xuzhou, and Jintan were 3.0, 6.7 and 10.3%, respectively (Table 1). These results indicated that the resistance of U. virens to DMIs evidently increased from 2015 to 2017. After isolating the six propiconazole-resistant isolates, we determined their resistance levels and investigated the influence of DMI fungicide resistance on their fitness. We found that the sensitive isolates couldn't grow on PSA plate with 0.4 μg/mL of propiconazole, but the resistant isolates could grow on PSA plate containing 10 μg/mL of propiconazole (Fig. 1). The RF values (Table 2) of these resistant isolates further demonstrated that the propiconazole-resistance levels of U. virens were low. In addition, the propiconazoleresistant isolates 2017-61 and 2017-170 did not reduce in mycelia growth, sporulation and pathogenicity compared to those of propiconazole-sensitive isolates, which indicated that their fitness values were high ( Table 4). The high fitness of propiconazole-resistant isolates suggested that the resistance of U. virens to DMI fungicides may continue to develop because of the strong competitiveness of resistant populations. Moreover, there was positive cross-resistance between propiconazole and other DMI fungicides (Table 3), indicating that the resistant isolates collected in the field were not selectively resistant to DMI fungicides at present.
As noted above, some pathogens are resistant to DMI fungicides. In most cases, resistance is linked to point Values are means ± SE of three biological replicates. ** P < 0.01 (t-test). c The level of CYP51 protein was assayed by western blot. The protein samples were respectively extracted from the propiconazole-sensitive and -resistant isolates. The concentration of propiconazole was 1 μg/mL mutations in the target gene CYP51 (De Waard 1996; Marichal et al. 1999;Wyand and Brown 2005;Cools and Fraaije 2013). The Y137H mutation in CYP51 protein in the resistant mutant of U. virens, previously reported to be generated by UV irradiation, conferred resistance to tebuconazole. In addition, overexpression of CYP51 was observed in the tebuconazole-resistant mutant . However, in the current study, no mutations in the CYP51 gene were detected in any of the six propiconazole-resistant isolates, which was consistent with the resistant isolates of U. virens screened in 2015 (Zhou et al. 2019).
In M. fructicola, a 65-bp sequence inserted at 113 bp upstream of the MfCYP51 gene was reported to be associated with DMI fungicide resistance (Chen et al. 2017). In this study, the 'CC' insertion was found in three resistant isolates (2017-61, 2017-170, and 2017-179) at 154 bp upstream of the CYP51 gene as previously reported in resistant isolate 88 collected in 2015 (Zhou et al. 2019). In addition, the expression levels of CYP51 and its encoded protein were significantly increased in these isolates (Fig. 2). The results suggested that 'CC' insertion mutation may influence the expression of CYP51 by altering the ability of transcription factors to participate in regulation, which may reduce sensitivity of U.
virens to DMI fungicides. In support of this inference, we used JASPAR (the largest transcription-factor binding profile database) to predict the functional binding sites of sequences with or without 'CC' insertion mutation (Wasserman and Sandelin 2004;Mathelier et al. 2014). As expected, 16 additional functional binding sites were found to be specific to the sequence with the 'CC' insertion mutation rather than to that without the 'CC' insertion mutation, and the binding abilities of two sites of sequence B were stronger than that of sequence A (Table 5).
In addition, most of the predicted transcription factors were involved in cell wall integrity or osmotic stress. For example, the Myb-like HTH transcription factor Dot6 in Candida albicans was reported to be involved in the TOR signaling pathway (Chaillot et al. 2019), which was regulated in the responses to cell wall damaging stress in F. graminearum (Gu et al. 2015). Hap2/5, CRZ1, ADR1, MIG1/2/3, MSN2/4, and RDS2 have been reported to function directly in cell wall integrity or osmotic stress (Proft and Serrano 1999;Mendizabal et al. 2001;Wong et al. 2003;Panadero et al. 2007  studies showing that the CYP51 gene encodes an enzyme that is part of the cell membrane, the current predictions further suggested that the resistance of U. virens to DMI fungicides may be related to CYP51 overexpression mediated by 'CC' insertion mutations. In our previous study, propiconazole stimulated the expression of CYP51 in both propiconazole-sensitive and -resistant isolates (Zhou et al. 2019). The same results were observed in this study (Fig. 2b). However, the level of CYP51-encoded protein was increased in propiconazolesensitive isolates and in propiconazole-resistant isolates that had no mutation with propiconazole treatment, whereas was unchanged in propiconazole-resistant isolates with the 'CC' insertion mutation (Fig. 2c). This result suggested that the DMI resistance may be induced by fungicides in the propiconazole-resistant isolates without any mutations. The underlying resistance mechanisms, for example, whether the ABC transporter is involved in these isolates, need to be further explored.

Conclusions
In this study, we showed a potential risk of U. virens resistance to DMI fungicides. Propiconazole-resistant isolates were found in three cities in Jiangsu Province. And all of them were positively cross-resistant to epoxiconazole, difenoconazole, and tebuconazole. In addition, two of them had high fitness. We suggest that it should be necessary to persistently monitor the resistance of U. virens to DMIs. DMIs should be valuable when used in alternation with other fungicides for the control of RFS in rice fields.

In vitro screening of propiconazole-resistant isolates
In order to obtain propiconazole-resistant isolates, mycelia plugs (5 mm in diameter) of 14-day-old isolates were transferred to PSA (potato sucrose agar: 200 g/L of potato, 20 g/L of sucrose, and 15 g/L of agar) plates containing 0.5 μg/mL propiconazole and then transferred to PSA plates with 1 μg/mL propiconazole and incubated at 28°C for 14 days. The isolates that grew on the PSA plates containing 1 μg/mL propiconazole were propiconazoleresistant isolates. EC 50 and EC 90 represented the concentration of propiconazole that resulted in 50 and 90% inhibition of U. virens. The average EC 50 value for the sensitive isolates was 0.05 μg/mL as Zhou et al. (2017) reported. The resistance factor (RF) in this study was the ratio of EC 50 value of resistant isolates to average EC 50 value (0.05 μg/ mL). RF values less than 10, between 10 and 100, and more than 100 were classified as low-resistance (LR), moderate-resistance (MR), and high-resistance (HR) strains, respectively.

Cross-resistance analysis
Cross resistance between propiconazole and other DMIs (epoxiconazole, difenoconazole, and tebuconazole) in U. virens was tested as previously described (Zhou et al. 2019). Mycelia plugs (5 mm in diameter) of each tested isolate were transferred to PSA plates containing different concentration of fungicides at 28°C for 16 days. For the propiconazole-sensitive isolates, the concentrations of propiconazole, epoxiconazole, difenoconazole, and tebuconazole were 0, 0.0125, 0.025, 0.05, 0.1, 0.2, and 0.4 μg/mL. For the propiconazole-resistant isolates, the concentrations of these fungicides were 0, 0.0195, 0.078, 0.3125, 1.25, 5, and 10 μg/mL. The 50 and 90% effective concentrations (EC 50 and EC 90 , respectively) of U. virens were calculated as described previously (Duan et al. 2018). There were three replicates of each concentration for each isolate. The experiment was repeated twice.

Mycelial growth, sporulation and pathogenicity
In order to test mycelial growth, mycelia plugs (5 mm in diameter) of 14-day-old isolates were transferred to PSA plates, and then the diameter of the colony was measured after incubation at 28°C for 18 days. In order to test sporulation, all tested isolates were cultured on PSA plates at 28°C for 14 days. Then two 5 mm-diameter mycelia plugs were transferred into 100 mL of PSB, and the conidium was counted with a hemocytometer after shaking at 28°C, 150 rpm for 7 days. In order to test virulence, the conidium and hyphal suspension was injected into rice panicles of Liangyoupeijiu as described by Yu et al. (2015). Each isolate was repeated three times and each experiment was conducted twice.

Sequence analysis of the CYP51 gene
Genomic DNA was extracted from the mycelia of all tested isolates using a DNA extraction kit (PD biotech, China). The primer pair Uv-CYP51-F/R listed in Additional file 1: Table S3 was designed to amplify the sequence of the CYP51 gene and its upstream region. PCR amplifications were performed as described previously (Zhou et al. 2019). The enzyme used in PCR reactions was Phanta Max Super-Fidelity DNA Polymerase (Vazyme, China). The PCR products were purified using Fastpure gel DNA extraction mini kit (Vazyme, China) and sequenced by Tsingke Biological Technology.

RT-qPCR and western blot assays
All the tested isolates were cultured in PSB liquid medium at 28°C for 4 days, and then treated with 1 μg/ mL of propiconazole for 24 h. Mycelia was harvested and washed with sterile water for total RNA and protein extraction. Total RNA and RT-qPCR were performed as described previously (Zhou et al. 2019). Briefly, RNA from mycelia of each isolate was extracted using a Total RNA Rapid Extraction Kit (BioTeke Co., Beijing, China), and reverse transcribed into cDNA using HiScript II 1st Strand cDNA Synthesis Kit (+gDNA wiper) (Vazyme Co., Nanjing, China) according to the manufacturer's instructions. RT-PCR was performed using primers listed in Additional file 1: Table S3. α-tubulin was used as a reference gene. Each isolate was repeated three times. For protein extraction, 50 mg of liquid nitrogen grinded mycelia was re-suspended in 500 μL of protein extraction buffer [50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 5 mM ethylene diamine tetraacetic acid (EDTA), 1% Triton X-100, 2 mM phenylmethylsulfonyl fluoride (PMSF)] supplemented with protease inhibitor cocktail (Beyotime, Shanghai, China). Each sample was loaded onto 12% SDS-PAGE gels. Anti-CYP51A1/CYP51 antibody (Abcam, Cambridge, England) and anti-beta actin antibody (Zoonbio, Nanjing, China) were used at 1:1000 for immunoblot analyses. After inoculation with secondary antibody anti-rabbit IgG-HRP (1:20,000) or anti-mouse IgG-HRP (1:20,000), chemiluminescence was detected as described previously (Li et al. 2019). In brief, the blots were determined by using an ECL substrate kit (Thermo Scientific, USA) and Tanon 5200 automatic chemiluminescent image analysis system (Tanon, Shanghai, China). The areas analyzed by Image J software was used to protein quantification.
Functional binding sites predicted using the JASPAR database The promoter of CYP51 was predicted using the online BDGP database (http://www.fruitfly.org/seq_tools/ promoter.html). Then, sequence A (a 93 bp nucleotide sequence of CYP51 promoter) and sequence B (a 95 bp nucleotide sequence of CYP51 promoter with the 'CC' insertion mutation) were used to predict functional binding sites in the JASPAR database (http:// jaspar.cgb.ki.se) ( Table 6). The relative profile score threshold was set to 80%.
Additional file 1: Table S1. Prediction of functional binding sites without 'CC' insertion sequences in the predicted promoter region of CYP51 in U. virens using the JASPAR database. Table S2. Prediction of functional binding sites with 'CC' insertion sequences in the predicted promoter region of CYP51 in U. virens using the JASPAR database. Table  S3. Primers used in this study.

Abbreviations
DMIs: Sterol demethylation inhibitors; EC 50 : The concentration of fungicides resulting in 50% inhibition of mycelia growth; EC 90 : The concentration of fungicides resulting in 90% inhibition of mycelia growth; PSA: Potato sucrose agar; RT-qPCR: Reverse transcription-quantitative PCR; RF: Resistance factors; RFS: Rice false smut Table 6 Nucleotide sequences used to predict functional binding sites in non-coding region upstream of CYP51 in Ustilaginoidea virens

Name
Nucleotide sequences sequence A gatctcgccttttgctcatcagcccaggcggagaggggaaccaagtcgaattgtataaatctggcagcggccgtcattcttctcgagccgaat sequence B gatctcgccttttgctcatcagcccCCaggcggagaggggaaccaagtcgaattgtataaatctggcagcggccgtcattcttctcgagccgaat The nucleotide sequence in bold font is the predicted promoter region, and the uppercase letters indicate the insertion mutation in propiconazole-resistant isolates