Introduction

China is one of the largest citrus-producing countries, ranked as the number one for the citrus-growing acreage and second to Brazil for the citrus yield in 2006. Green mold caused by Penicillium digitatum (Pers.: Fr.) Sacc. is the most destructive postharvest disease of citrus. It is responsible for about 90% of the losses during postharvest citrus packing, storage, transporting, and marketing (Kanetis et al. 2007; Macarisin et al. 2007). Immediate application of fungicides after harvesting is very effective and used worldwide for green mold control (Smilanick et al. 2005, 2006). In China, benzimidazoles, such as carbendazim, thiabendazole, and thiophanate methyl had been used for green mold control before the mid-1990s. They were replaced gradually by imazalil (IMZ) and prochloraz due to the widespread emergence of benzimidazole resistance in P. digitatum (Jiang et al. 2010; Li et al. 2003; Zhu et al. 2006). Imazalil and prochloraz are currently the primary fungicides for postharvest green mold control of citrus in China.

Both imazalil and prochloraz belong to the demethylation inhibitors (DMIs) of ergosterol biosynthesis. They inhibit cytochrome P450-dependent sterol 14α-demethylase (P45014αdm) activity, following the stoichiometric interaction of the N-3 (imidazoles) or the N-4 (triazoles) substituents of the azole ring with the heme iron of the P45014αdm (Yoshida and Aoyama 1987; Yoshida 1988). Because of its different mechanism of action, imazalil is effective against P. digitatum resistant to benzimidazole. Imazalil has been used for postharvest citrus disease control since 1976 (Harding 1976). As a consequence of continuous use, packing house imazalil-resistant isolates have occurred in China and elsewhere (Li et al. 2003; Chen et al. 2008; Rebellato and Monteiro 1984; Eckert 1987; Bus et al. 1991; Wild 1994; Sánchez-Torres and Tuset 2011), and the control efficiency has been compromised (Eckert et al. 1994; Jiang et al. 2010).

Previously, two similar types of mutations have been found to be associated with imazalil resistance of P. digitatum. These are: (1) the insertion of four extra tandem copies of a 126-bp transcriptional enhancer in the promoter region of the CYP51A gene, which resulted in the overexpression of CYP51A (IMZ-R1; Hamamoto et al. 2000); and (2) the insertion of a 199-bp sequence within the 126-bp transcriptional enhancer, which also resulted in the elevated expression of CYP51A (IMZ-R2; Ghosoph et al. 2007). In addition, overexpression of the ATP-binding cassette efflux transporters, PMR1 and PMR5, was also shown to contribute to the resistance to imazalil and other fungicides in P. digitatum (Nakaune et al. 1998, 2002).

The resistance of P. digitatum to DMIs in Spain was evaluated recently (Sánchez-Torres and Tuset 2011). Regarding CYP51 gene, only IMZ-R1 genotype was found in three of 58 IMZ-R isolates. Several mutations in promoter and the coding region of PMR5 were found to be correlated with DMI fungicide resistance in IMZ-R isolates compared with IMZ-S ones. They also proved the PMR1 gene might play a partial role in determining the baseline resistance to DMIs. However, more IMZ-R isolates indicated that other genes more than CYP51 and PMR5 could involve in DMI fungicide resistance.

In our previous study, both IMZ-R1 (Pd01) and IMZ-R2 (Pd19d) of P. digitatum were detected in Zhejiang province, China (Li et al. 2003; Zhu et al. 2006; Chen et al. 2008). However, further investigation indicated that the mechanism underlying imazalil resistance in many of the imazalil-resistant isolates collected in Zhejiang, especially from Quzhou, was neither IMZ-R1 nor IMR-R2, and the expression levels of PdCYP51A in these isolates were comparable to that of the IMZ-S ones examined. Additionally, point mutations associated with DMI resistance, as reported in other phytopathogenic fungi (Delye et al. 1997, 1998; Cools et al. 2002; Albertini et al. 2003; Fraaije et al. 2007; Leroux et al. 2007; Canas-Gutierrez et al. 2009; Stammler et al. 2009), were not detected in the promoter and the coding region of PdCYP51A. Thus, we concluded that a novel imazalil resistance mechanism, termed IMZ-R3, must be present in these resistant isolates.

Recently, the homologous CYP51 genes found in some filamentous fungi were approved to contribute to the sensitivity to DMI fungicides in Aspergillus fumigatus (Martel et al. 2010) and Fusarium graminearum (Liu et al. 2011). Based on these studies, we cloned two homologous CYP51 genes in P. digitatum, designated PdCYP51B and PdCYP51C. Further experiments indicated that a 199-bp insertion in the promoter region of the PdCYP51B gene that resulted in the overexpression of PdCYP51B was responsible for the observed resistance mechanism of IMZ-R3 isolates of P. digitatum to imazalil.

Materials and methods

Fungal isolates and cultivation

The monoconidial isolates used in this study were isolated from P. digitatum-infected citrus fruits collected from packing houses in Quzhou, Jinhua, Lishui, and Hangzhou of Zhejiang province, China, between 2000 and 2010, and were maintained on potato dextrose agar medium (PDA) at 4°C. Mycelium of P. digitatum was cultured in liquid potato dextrose on a rotary shaker (160 rpm) at 25°C, whereas the conidial masses were harvested from colonies cultured on PDA at 25°C.

Determination of IMZ-S or IMZ-R phenotypes and resistance mechanisms of P. digitatum

The phenotype (imazalil-sensitive or imazalil-resistant) of the collected P. digitatum isolates was determined by growing them on imazalil-amended PDA as described previously (Ghosoph et al. 2007; Holmes and Eckert 1999). IMZ-S isolates were showing no or little growth after 5 days at 25°C on PDA amended with 0.1 μg/ml (or less) of IMZ. IMZ-R isolates were capable of growing after 5 days at 25°C on PDA with 0.5 μg/ml or more of IMZ. Each isolate was tested twice for IMZ sensitivity.

The genotypes or molecular mechanisms of imazalil resistance were determined using primer pair CYP51A1/CYP51A2, designed specifically for the detection of IMZ-R1 (insertion of four extra tandem repeats of the 126-bp transcriptional enhancer in the promoter region of PdCYP51A) and IMZ-R2 (199-bp insertion mutation in the promoter region of PdCYP51A gene), respectively.

Oligonucleotide primers

The oligonucleotide primers used in this study are listed in Table 1. The positions of these primers within and around the coding region of the PdCYP51B gene are shown in Fig. 1.

Table 1 Primers used in this study
Full size table
Fig. 1
figure 1

Schematic diagram of the promoter and coding region of the PdCYP51B gene. The blank arrow represents the 199-bp insertion in the promoter region of the PdCYP51B gene. The dotted box represents the coding region of the PdCYP51B which contains four exons indicated by the solid arrows and three introns indicated by solid lines between the exons. The positions of selected primers are indicated in the diagrams

Full size image

Cloning of PdCYP51B and PdCYP51C

Based on P. digitatum transcriptome data (unpublished), we found two EST sequences, homologous to the formerly reported CYP51 (PdCYP51A; GenBank, AJ439080.1). They were designated as PdCYP51B and PdCYP51C. According to the Penicillium chrysogenum genome sequence (NCBI genome project, NS_000201), the primer pairs CYP51B-F/CYP51B-R and CYP51C-F/CYP51C-R were used for amplifying the genomic DNA fragment, including the complete coding region and part of the upstream and downstream region sequences of PdCYP51B and PdCYP51C. The amplified fragments were cloned into the PMD18-T vector (TaKaRa Biotech. Co., Dalian, China) and sent for sequencing.

To determine the gene structure of PdCYP51B and PdCYP51C, total RNA was extracted using the AxyPrepTM multisource total RNA miniprep Kit (Axygen, USA). First-strand cDNA was synthesized using the RNA PCR Kit (AMV) 3.0 kit (TaKaRa Biotech. Co.). The full-length PdCYP51B and PdCYP51C genes were amplified from the cDNA template, using primer pairs B-CDS-F/B-CDS-R and C-CDS-F/C-CDS-R.

Sequence analysis of PdCYP51B and PdCYP51C

PdCYP51B and PdCYP51C and their partial upstream and downstream regions were amplified, using primers pairs CYP51B-F/CYP51B-R and CYP51C-F/CYP51C-R, from eight IMZ-S isolates, six IMZ-R3 isolates, one IMZ-R1 isolate, and one IMZ-R2 isolate of P. digitatum. The obtained sequences were aligned by ClustalX 2.0 (http://www.clustal.org) for identification of DNA differences. The NSITE program (www.softberry.com) and the eukaryotic promoter predictor (Berkeley Drosophila Genome Project, http://www.fruitfly.org/seq_tools/promoter.html) were used to analyze the sequence of PdCYP51B gene. The protein sequences obtained of genes PdCYP51A, PdCYP51B, and PdCYP51C and of related fungal and plant proteins obtained from the NCBI GenBank were aligned using ClustalX 2.0, and phylogenetic analysis of the protein sequences was conducted by MEGA 4.1 (http://www.megasoftware.net).

Gene expression analysis by real-time RT-PCR

The relative expression levels of PdCYP51A, -B, and -C of selected P. digitatum isolates with different levels of imazalil resistance were assayed by real-time reverse transcription PCR (RT-PCR) on a 7300 Real-Time PCR system (ABI, USA). Total RNA extraction and first-strand cDNA synthesis were described previously (Wang and Li 2008). Real-time PCR was carried out using the SYBR Premix Ex Taq TM (Perfect Real Time) kit (TaKaRa Biotech. Co.). The thermal cycling conditions were 95°C for 30 s, 40 cycles of 95°C for 5 s, and 60°C for 31 s. The primer pairs qCYP51A-F/qCYP51A-R, qCYP51B-F/qCYP51B-R, and qCYP51C-F/qCYP51C-R (Table 1) were designed for relative expression analysis of PdCYP51A, PdCYP51B, and PdCYP51C, respectively. The P. digitatum actin-encoding gene (GenBank, AB030227), amplified with primers q2-actin-R/q2-actin-F, was used to normalize the quantification of gene expression.

In all tests, negative controls containing no template cDNA were subjected to the same procedure to eliminate or identify any possible contamination. All genes were amplified in three separate reactions using the same cDNA template, and the mean values were obtained. Each experiment was conducted in triplicate. In the validation experiment, the log cDNA dilution of ten-fold serially diluted cDNA aliquots versus C t, was used to calculate PCR efficiency of each gene according to the equation E = 10(−1/slope). The relative quantification of the target gene in comparison with the reference actin gene was calculated according to the formula, \( {\text{Ratio}} = \left[ {{{\left( {{E_{\text{target}}}} \right)}^{{\Delta {C_{\text{t}}}{\text{ target }}({\text{control}} - {\text{sample}})}}}\left] / \right[{{\left( {{E_{\text{reference}}}} \right)}^{{\Delta {C_{\text{t}}}{\text{ reference }}({\text{control}} - {\text{sample}})}}}} \right] \) (Pfaffl et al. 2002). ANOVA was applied to determine significant differences among the C t of the examined isolates (Tang and Feng 2007).

Induced expression of PdCYP51A, PdCYP51B, and PdCYP51C of P. digitatum was studied in IMZ-S isolate PdKH8 by adding imazalil to a final concentration of 0.1 μg/ml to 4-day-old mycelial cultures of P. digitatum grown in liquid potato medium. After addition of imazalil, cultures were incubated at 25°C, 160 rpm for 6 h. Then, the mycelia were harvested by filtrating through a two-layer cheesecloth, washed with ddH2O, frozen in liquid nitrogen, and stored at −80°C until RNA isolation.

Construction of the PdCYP51B overexpression plasmid

The full-length PdCYP51B as well as its upstream fragment was amplified from IMZ-R (PdW03) and IMZ-S (PdKH8) isolates using primer pair CYP51B-EX-R/CYP51B-EX-F, into which 5′-terminal SacI and KpnI restriction sites were introduced. The PCR products were digested with SacI and KpnI and cloned into the pTFCM to generate pTFCM-PdCYP51B (Fig. 2a).

Fig. 2
figure 2

Construction and identification of PdCYP51B overexpression mutants. a Diagram for construction of the PdCYP51B overexpression plasmid. PdCYP51B S and PdCYP51B R were inserted into SacI and KpnI sites of plasmid pTFCM, respectively, to produce the corresponding pTFCM-PdCYP51B S and pTFCM- PdCYP51B R constructs. b PCR identification with primers B1 and B2 of PdCYP51B overexpression mutants OE-PdCYP51B S and OE-PdCYP51B R. c Southern blot analysis of OE-PdCYP51B S and OE-PdCYP51B R. The fungal DNA was digested by KpnI, and hph (hygromycin-resistant gene) was used as the probe

Full size image

P. digitatum transformation and analysis of transformants

The imazalil-sensitive isolate PdKH8 was used for Agrobacterium tumefaciens-mediated transformation (ATMT) with the A. tumefaciens strain AGL-1, as described previously (Wang and Li 2008), to create PdCYP51B overexpression mutants, called OE-PdCYP51B R and OE-PdCYP51B S. OE-PdCYP51B R and OE-PdCYP51B S were confirmed by PCR using primers B1/B2 (Table 1, Fig. 2b). Southern blot (Fig. 2c) was carried out as described previously (Wang and Li 2008). Briefly, genomic DNA (5–30 μg) of individual isolates was digested with KpnI. Part of the hygromycin-resistant gene hph was used as a hybridization probe and labeled with digoxigenin using the DIG High Primer DNA Labeling and Detection Starter Kit II (Roche, Mannheim, Germany) according to the manufacturer's instructions.

DMI resistance assays

The resistant levels of mutants OE-PdCYP51B S (its exogenous PdCYP51B was derived from an IMZ-S isolate), OE-PdCYP51B R (its exogenous PdCYP51B derived from an IMZ-R3 isolate), and the parental isolate PdKH8 (imazalil-sensitive) to imazalil and other DMIs were assayed on PDA containing imazalil and other three DMIs, as described previously. The following fungicide concentrations were used at 0.04, 0.1, and 1 μg/ml for imazalil; 0.02, 0.05, and 0.2 μg/ml for difenoconazole; 0.2, 0.3, and 0.7 μg/ml for tebuconazole; and 0.3, 0.5, and 2.5 μg/ml for myclobutanil. The assays were conducted three times.

PCR-based rapid detection of imazalil resistance in P. digitatum

To establish a rapid molecular method to detect imazalil resistance of P. digitatum and their genotypes of resistance, the primer pair B1/B2 was designed (Table 1). By combining primers CYP51A1/CYP51A2 (Table 1) that were used to distinguish the IMZ-R1 and IMZ-R2 (Chen et al. 2008), a duplex PCR was developed to detect imazalil-resistant genotypes. Rapid DNA extraction was conducted according to the method described by Chen et al. (2008). PCR was performed with a total volume of 20 μl containing 2 μl Taq Buffer (Mg2+ plus), 2 μl dNTP (2.5 mM), 1 μl of each of four primers (5 μM), 0.1–1 μg DNA template, 0.2 μl Taq polymerase (5 U/μl), and distilled water up to 20 μl. A gradient PCR with thermal cycling conditions 95°C 3 min, 32 cycles of 95°C 40 s, 55–62°C 40 s, and 72°C 1 min was conducted to confirm the optimal annealing temperature of the mixed primers.

Results

Sensitivity to imazalil and resistance mechanisms of P. digitatum

In total, 403 isolates of P. digitatum collected during 2000 to 2010 were tested for their IMZ-S or IMZ-R phenotypes (Table 2). One hundred twenty isolates were from 15 supermarkets in Hangzhou in 2000; 12 isolates were from four packing houses in Jinhua in 2005, 6 isolates from two packing houses in Lishui in 2005, and the other 265 isolates were from 22 packing houses in Quzhou during 2000 to 2010, the main citrus-producing region in Zhejiang. No sequential year sampling was taken in Quzhou. The result showed that IMZ-R1 isolate was firstly detected Quzhou and Hangzhou in 2000 (Zhu et al. 2006), then detected again in Quzhou in 2008, accounting for 6.5% of the total IMZ-R. IMZ-R2 isolates were firstly found in Jinhua in 2005, then in Quzhou in 2008, representing 4.5% of the total IMZ-R, whereas the unknown resistant genotype termed as IMZ-R3 was predominant (89%) in the IMZ-R population (Table 2).

Table 2 Isolates of P. digitatum collected from Zhejiang during 2000 to 2010
Full size table

Cloning and sequencing of PdCYP51B and PdCYP51C

To explore whether any homologous genes of CYP51, as reported in A. fumigatus (Mellado et al. 2001), were present in P. digitatum, we manually checked our annotated P. digitatum transcriptome database (unpublished) and found two EST sequences which were designated CYP51B and CYP51C. To obtain the full as well as the partial sequences upstream and downstream of the PdCYP51B and PdCYP51C genes, primer pairs of CYP51B-R/CYP51B-F and CYP51C-R/CYP51C-F were designed according to the corresponding sequences in P. chrysogenum. Results indicated that the cloned fragment with CYP51B-R/CYP51B-F from IMZ-S isolate PdKH8 contained 2,919 bp, encompassing a complete coding region (1,751 bp) and its upstream (742 bp) and downstream (425 bp) sequences. RT-PCR using primers B-CDS-F/B-CDS-R verified that this gene contained an opening reading frame (ORF) of 1,751 bp, with three introns of 73, 51, and 52 bp, located between positions 989–1,061; 1,260–1,310; and 2,376–2,427 bp, respectively (Fig. 1).

PdCYP51C and partial sequences of its upstream and downstream were obtained using the same method. The cloned fragment with CYP51C-R/CYP51C-F had 2295 nt, containing an ORF of 1,497 bp, with five introns located between 541–596, 682–733, 823–885, 1,177–1,242, and 1,292–1,348 bp, respectively (data not shown). The nucleotide sequences of PdCYP51B and PdCYP51C genes were deposited in the GenBank with accession number HQ724322 and HQ724324.

Phylogenetic analysis (Fig. 3) showed that the deduced amino acid sequence of PdCYP51B was 78% and 61% identical to AfCYP51B of A. fumigatus (GenBank, XP_749134.1) and FgCYP51B of F. graminearum (Broad Institute, FGSG_01000), respectively. The PdCYP51C showed 19% identity to FgCYP51C of F. graminearum (FGSG_11024) and FoCYP51C of Fusarium oxysporum (FOXG_13138). Moreover, PdCYP51B was 59% and 22% identical to PdCYP51A and PdCYP51C, respectively, while PdCYP51C was 23% identical to PdCYP51A (data not shown). The percentage of identity at the amino acid level was high enough to consider PdCYP51B a member of the fungal CYP51 family.

Fig. 3
figure 3

Phylogenetic tree of PdCYP51A, PdCYP51B, and PdCYP51C with other CYP51s. Phylogenetic tree was generated by the bootstrapped neighbor-joining method with Mega 4.1 software on the basis of deduced amino acid sequences of PdCYP51A (BAB03658.1), PdCYP51B (HQ724322), and PdCYP51C (HQ724324) from P. digitatum isolate KH8 and those from fungal species A. fumigatus (AfCYP51A, XP_752137.1; AfCYP51B, XP_749134.1), Monilinia fructicola (MfCYP51, ACY41226.1), Mycosphaerella graminicola (MgCYP51, AAP79601.1), F. graminearum (FgCYP51A, FGSG_04092, FgCYP51B, FGSG_01000, FgCYP51C, FGSG_11024), F. oxysporum (FoCYP51C, FOXG_13138), Blumeria graminis (BgCYP51, CAE17515.1), C. albicans (CaCYP51, BAB03399.1), Botryotinia fuckeliana (BfCYP51, AAK26391.1), S. cerevisiae (ScCYP51, DAA06695.1), Ustilago maydis (UmCYP51, CAA88176.1), Arabidopsis thaliana (AtP450, AAD30262.1), Venturia inaequalis (ViCYP51A, AAF76464.1), and M. oryzae (MoCYP51A, MGG_04628.6; MoCYP51B, MGG_04432.6). The CYP51s from P. digitatum are indicated in boxes

Full size image

Presence of a unique 199-bp insertion in the promoter region of PdCYP51B in IMZ-R3 of P. digitatum

PdCYP51B and PdCYP51C were amplified from eight IMZ-S isolates (PdHL5, PdHL6, PdKH1, PdJ1, PdYX1, PdYX2, PdW07, and PdKH8), one IMZ-R1 (Pd01) isolate, one IMZ-R2 (Pd19d) isolate, and six IMZ-R3 isolates (PdXJ2, PdDQ8, PdXJ3, PdW03, PdQZ15, and PdXJ8) and sequenced. Alignment of nucleotide sequences of these PdCYP51B genes showed that an extra 199-bp fragment insertion was present at the position −174 bp (from ATG) of the PdCYP51B promoter region in all six IMZ-R3 isolates but was absent in both IMZ-R1 and IMZ-R2, as well as in all eight IMZ-S isolates (Figs. 1 and 4). Apart from this difference, no point mutations that related to imazalil resistance were detected either in the encoding or in the promoter regions of PdCYP51B of P. digitatum (data not shown).

Fig. 4
figure 4

Analysis of the 199-bp insert. Transcription factor binding sites within the 199-bp sequence found by means of a computer search on the regulatory element database using the NSITE program (www.softberry.com). Promoter analysis was performed by eukaryotic promoter predictor (Berkeley Drosophila Genome Project, http://www.fruitfly.org/seq_tools/promoter.html). The sites found in the upper strand are shown above the sequence, and the sites in the lower strand are shown below the sequence. The box shows the putative promoter region. The TSD and imperfect TIR are indicated by the solid and blank arrows, respectively

Full size image

PdCYP51C was also amplified from the above 16 isolates. However, neither an insertion mutation nor a consistent point mutation was detected in the coding or the promoter regions that correlated with imazalil resistance (data not shown).

To test if the 199-bp insertion in the promoter region of PdCYP51B was universal in all collected IMZ-R3 isolates of P. digitatum, the primer pair B1/B2 (Table 2) was used for PCR amplification from the following P. digitatum isolates: 36 IMZ-R3, 3 IMZ-R1, 3 IMZ-R2, and 20 IMZ-S. As expected, a 600-bp fragment was obtained from all 36 IMZ-R3, while a 400-bp fragment was amplified from the other isolates of P. digitatum, indicating that this 199-bp insertion was universal in IMZ-R3, and it must be correlated with imazalil resistance.

Bioinformatical analysis of the 199-bp insertion sequence

Analyzing the 199-bp insertion sequence (Fig. 4), we found a 6-bp target site duplication (TSD) and the imperfect terminus inverted-repeat (TIR) sequences. Moreover, four sites for known fungal transcriptional binding factors (Aspergillus nidulans abaA, Neurospora crassa CYS-3, Saccharomyces cerevisiae Reb1, and S. cerevisiae CLN3) and 11 sites for different vertebrate binding factors (data not shown) were predicted within the 199-bp insertion by NSITE program (Fig. 4). A putative promoter region was also found by the eukaryotic promoter predictor (Fig. 4).

Expression of PdCYP51B is constitutively higher in IMZ-R3 isolates of P. digitatum

To explore if the 199-bp insertion changed the expression of the PdCYP51B gene, the relative expression levels of PdCYP51B of three IMZ-R3 isolates (PdW03, PdXJ8, and PdDQ8), two IMZ-S isolates (PdKH8 and PdKH1), one IMZ-R1 isolate (Pd01), and one IMZ-R2 isolate (Pd19d) were determined by qPCR. The results showed that the transcription levels of PdCYP51B in three IMZ-R3 isolates were about 7.5–13.6 times higher than that of two IMZ-S isolates (Table 3, P < 0.01). However, no significant expression differences of PdCYP51A and PdCYP51C were observed among these three IMZ-R3 isolates and two IMZ-S isolates tested (Table 3), indicating that constitutive overexpression of the PdCYP51B gene was correlated with the presence of 199 bp insertion in the promoter region of PdCYP51B. Additionally, the expression level of PdCYP51A instead of PdCYP51B and PdCYP51C in Pd01 and Pd19d was significantly higher than that of IMZ-S and IMZ-R3 of P. digitatum (Table 3), consistent with previous reports by Hamamoto et al. (2000) and Ghosoph et al. (2007).

Table 3 Relative expression of the PdCYP51A, PdCYP51B, and PdCYP51C genes
Full size table

PdCYP51A, PdCYP51B, and PdCYP51C are inducible by imazalil

To examine if the expression of PdCYP51A, PdCYP51B, and PdCYP51C was inducible by imazalil, the expression levels of these genes pre- and posttreated (6 h) with 0.1 μg/ml imazalil were compared in the imazalil-sensitive isolate PdKH8. The results showed that all of these genes were significantly upregulated after imazalil treatment (Fig. 5, P < 0.05).

Fig. 5
figure 5

Relative expression levels of PdCYP51A, PdCYP51B, and PdCYP51C in P. digitatum after treatment with imazalil. The IMZ-S isolate PdKH8 was treated with 0.1 μg/ml imazalil for 6 h. The relative expression of PdCYP51A, PdCYP51B, and PdCYP51C in treated and untreated PdKH8 were compared

Full size image

Overexpression of PdCYP51B leads to reduced sensitivity of P. digitatum to imazalil

To determine if overexpression of PdCYP51B could lead to reduced sensitivity of P. digitatum to imazalil, the full-length PdCYP51B, including the complete coding region and the promoter region, was amplified from PdKH8 (termed PdCYP51B S) and PdW03 (termed PdCYP51B R) using primers CYP51B-EX-R/CYP51B-EX-F and introduced into PdKH8 by ATMT. Selected transformants were verified by PCR with primers B1/B2. For OE-PdCYP51BS transformants, one 401-bp fragment was amplified, whereas, in the OE-PdCYP51B R transformant, two fragments of 401 and 600 bp were amplified (Fig. 2b). Southern blot using portion of the hph gene (626 bp) as a probe demonstrated that a single target gene insertion occurred in both OE-PdCYP51B S and OE-PdCYP51B R.

Resistance of OE-PdCYP51B S and OE-PdCYP51B R mutants, as well as the parental isolate PdKH8 of P. digitatum to DMI fungicides, was compared on PDA containing different concentrations of imazalil, difenoconazole, tebuconazole, and myclobutanil, respectively. The results showed that both PdCYP51B S and PdCYP51B R confer PdKH8 with increased DMI fungicide resistance; however, the effect of PdCYP51B R was significantly higher than that of PdCYP51B S (Fig. 6), indicating that expression levels of PdCYP51B had a major role in determining the sensitivity of P. digitatum to DMI fungicides and that the insertion of 199 bp in the promoter region of PdCYP51B indeed significantly increased the resistance of P. digitatum to imazalil and other DMI fungicides.

Fig. 6
figure 6

Comparisons of sensitivity to different DMI fungicides among mutants of OE-PdCYP51B and their parental isolate PdKH8 of P. digitatum. OE-PdCYP51B S, the overexpression mutant of the PdCYP51B gene amplified from IMZ-S P. digitatum; OE-PdCYP51B R, the overexpression mutant of the PdCYP51B gene amplified from IMZ-R3 P. digitatum; PdKH8, the parental isolate. Concentrations of DMI fungicides are indicated on the top of each plate, whereas the fungicides are presented in left column

Full size image

PCR-based rapid detection of DMI-resistant phenotypes and their resistant genotypes in P. digitatum

Based on the discovered molecular mechanisms of DMI resistance of P. digitatum, a PCR-based rapid detection for DMI-resistant genotypes was developed by using primer pair CYP51A1/CYP51A2 (Chen et al. 2008) together with B1/B2 (this study). Optimization of annealing temperature through gradient PCR indicated that 56°C was the ideal annealing temperature to obtain the best amplification in terms of specificity and efficiency. With these pairs of primers, specific fragments of 401 and 506 bp were generated from DMI-sensitive isolates (Fig. 7, lines 4 and 5); 401 and 1,010-bp fragments were amplified from IMZ-R1 isolates (Fig. 7, lines 1 and 5); 401 and 705-bp fragments were obtained from IMZ-R2 isolates (Fig. 7, lines 2 and 5); and fragments of 506 and 600 bp were amplified from IMZ-R3 isolates (Fig. 7, lines 3 and 4).

Fig. 7
figure 7

PCR-based rapid detection of DMI resistance in P. digitatum strains. PCR was performed using primers CYP51A1/CYP51A2 and B1/B2 with the annealing temperature of 56°C

Full size image

Discussion

Ergosterol is an essential component of fungal cell membranes, serving the same function as cholesterol in animal cells (Espenshade and Hughes 2007). Given the absence of ergosterol in mammals, ergosterol biosynthesis pathway is an ideal target for antifungal drugs development (Georgopapadakou and Walsh 1994). Sterol 14α-demethylase (CYP51) is a key enzyme in the sterol biosynthesis pathway in fungi (Mitropoulos et al. 1976; Trzaskos et al. 1986). Thus far, more than 400 redundant CYP51 amino acid sequences from 20 genera in fungi are available in GenBank. The copies of CYP51 in fungi vary depending on the species. In S. cerevisiae and Candida albicans, only one erg11/CYP51 gene was found (Ferreira et al. 2005). However, in filamentous fungi, more than one CYP51 genes were usually identified, e.g., two in A. fumigatus and Magnaporthe oryzae (Ferreira et al. 2005; Yan et al. 2011), three in Aspergillus oryzae, F. graminearum, and F. oxysporum (Ferreira et al. 2005; Yin et al. 2009; Liu et al. 2011).

Although homologous genes of CYP51 have been found in some fungal species, the precise roles of these genes, especially the CYP51B gene, in mediating fungicide resistance have been poorly understood. In A. fumigatus, point mutations of AfCYP51A amino acid at different positions have been reported to be correlated with azole resistance, but resistance mechanisms related to AfCYP51B gene were not detected in naturally occurred azole-resistant isolates (Verweij et al. 2009). However, heterologous expression of AfCYP51A and AfCYP51B in an S. cerevisiae mutant (YUG37-erg11) resulted in a similar increase of resistance to four azole fungicides tested, indicating that both AfCYP51A and AfCYP51B are involved in azole resistance (Martel et al. 2010). Deletion of FgCYP51A in F. graminearum resulted in increased sensitivity to seven DMI fungicides (triadimefon, propiconazole, tebuconazole, diniconazole, difenocoazole, flutriafol, and prochloraz) studied, and the deletion of FgCYP51C resulted in increased sensitivity to five (tebuconazole, diniconazole, difenocoazole, flutriafol, and prochloraz) of the seven DMIs mentioned above. However, deletion of FgCYP51B did not change F. graminearum's sensitivity to these DMIs, indicating that FgCYP51B may not have a role in DMI resistance in F. graminearum (Liu et al. 2011). Similar results were shown for CYP51B in M. oryzae (Yan et al. 2011).

In this study, two genes (PdCYP51B and PdCYP51C) homologous to PdCYP51A were cloned in P. digitatum. Both PdCYP51B and PdCYP51C were inducible by imazalil treatment (Fig. 5). Introduction of another copy of PdCYP51B S did increase the resistance of P. digitatum to DMIs (Fig. 6). These results definitely demonstrated that PdCYP51B is involved in DMI resistance for P. digitatum. Quantitative PCR showed that constitutive expression levels of PdCYP51B in IMZ-R3 P. digitatum were significantly higher than that of IMZ-S as well as of IMZ-R1 and IMZ-R2 (Table 3). A unique 199-bp insertion in the promoter region of PdCYP51B was found to be responsible for the increased expression of PdCYP51B. The positive effect of 199-bp insertion on DMI resistance was then supported by the genetic evidence since OE-PdCYP51B R showed significantly higher resistance to DMIs than that of OE-PdCYP51B S (Fig. 6). Thus, it could be concluded that both PdCYP51A and PdCYP51B were involved in DMI resistance, and the increased expression of PdCYP51B resulted from the 199-bp nucleotide-insertion mutation in the promoter region of PdCYP51B was responsible for IMZ-R3, the predominant IMZ-R genotype of P. digitatum in Zhejiang, China. To our best knowledge, this is a new mechanism of DMI resistance for fungi.

Since PdCYP51B and PdCYP51C were initially identified from the transcriptome database of P. digitatum (unpublished), both genes should be expressed and thus were unlikely to be pseudogenes. Sequence alignment indicated that PdCYP51B was ≥78% identical to CYP51B from A. fumigatus (XP_749134.1). Although PdCYP51C was only 18–23% identical to other CYP51s of fungi, it was still assigned to the CYP51 family when BLAST searching in GenBank database.

Analysis of this 199-bp insert showed that no typical signs of a transposon (e.g., transposase-encoded sequence) were found, but a 6-bp TSD and the imperfect TIR were presented at both terminus of the fragment, indicating that the 199-bp insert could be a miniature inverted-repeat transposable element (MITE, Fig. 4). A promoter located inside this 199-bp fragment was predicted by bioinformatical tools, but it was absent in the putative promoter region of PdCYP51B of IMZ-sensitive P. digitatum. In addition, four known fungal transcription factors binding sites were found in the 199-bp fragment. Thus, an unproved but reasonable inference is that this putative MITE acting as both a transcription enhancer and a promoter inserts into the upstream of PdCYP51B gene, and such insertion event leads to the increased expression of PdCYP51B, therefore the reduced sensitivity of P. digitatum to DMI fungicides. P. digitatum bearing this mutation is selected, the population of this genotype increases, then becomes predominant under sustained DMI fungicide pressure introduced by the common usage of imazalil, one of DMIs, for postharvest disease control of citrus since the middle 1990s in Zhejiang province (Li et al. 2003; Chen et al. 2008), particularly in Quzhou, where the IMZ-R3 isolates were obtained for this study.

The origin of IMZ-R1 and IMZ-R2 P. digitatum in California were determined by RAPD analysis (Ghosoph et al. 2007), the result indicated that imazalil resistance developed independently. Together with our study on IMZ-R P. digitatum in Zhejiang, it could be concluded that P. digitatum has simultaneously evolved multiple mechanisms to adapt to the selection pressure introduced by the application of DMI fungicides in modern agriculture.

The application of synthetic fungicides is the primary method in citrus postharvest disease control. The evolution of fungicide-resistant pathogens becomes a significant problem for citrus postharvest disease control because control systems are less effective or no longer effective (Bus et al. 1991; Eckert et al. 1994). The detection and quantification of fungicide-resistant pathogenic populations would be useful in assisting to determine whether the ineffective disease control results from fungicide resistance. The most common method to determine the fungicide resistance of a pathogen is based on the inhibition of mycelial growth sensitivities measured by either EC50 values or minimal inhibitory concentrations. Molecular detection and quantification of fungicide-resistant populations have been demonstrated to be an alternative and effective way in diagnosis of resistant populations (Ma et al. 2005; Chen et al. 2008; Ghosoph et al. 2007; Zhang et al. 2009). In the current study, our work not only extends the knowledge of DMI-resistant mechanism in fungi but also provides a foundation to develop a practical method for rapidly identifying and monitoring the incidence of DMI-resistant isolates among populations of conidia within citrus packing houses, an important commercial practice for citrus.