Colletotrichum gloeosporioides was isolated from the H. brasiliensis with Anthracnose; furthermore, the genome of the C. gloeosporioides was sequenced for exploration of possible mechanism of pathogenicity (data unpublished). The C. gloeosporioides was used as recipient strain for the transformation experiments and as a WT control. All the C. gloeosporioides strains were grown on potato dextrose agar (PDA) at 28°C.

Vector pCB1532 carrying the acetolactate synthase gene (SUR) cassette from M. oryzae (Yang et al., 2013) conferred resistance to chlorimuron ethyl (a sulfonylurea herbicide) and pKOV21 carrying the hygromycin phosphotransferase gene (HPH) conferred resistance to hygromycin were used to construct the replacement vectors. The replacement vector for Dcl1 was designed as described in Figure 1A: the 5′ flanking region and 3′ region of the Dcl1 nucleotide were amplified by use of the primer pairs 1/2 and 3/4, respectively; then the two fragments were excised with EcoRI/HindIII and XbaI/EcoRI respectively and ligated into vector pCB1532 to construct the replacement vector. The replacement vector for Dcl2 was designed as described in Figure 1B: the 5′ flanking region and 3′ region of the Dcl2 nucleotide were amplified by use of the primer pairs 5/6 and 7/8, respectively; then the two fragments were excised with EcoRI/HindIII and XbaI/EcoRI respectively and ligated into vector pKOV21 to construct the replacement vector. The two deletion vectors were excised with EcoRI before transformation of the C. gloeosporioides strain. Vector pBS-NEO carrying the Neomycin phosphotransferase gene (NPTII) conferred resistance to Geneticin (G418) was used to construct the complementation vector. To get the complementation vector, a 6.4 kb fragment containing Dcl1 the sequence together with a 1.4 kb upstream nucleotide and a 5.9 kb fragment containing Dcl2 the sequence together with a 1.3 kb upstream nucleotide were amplified by using the primers 21/22 and 23/24; then the two fragments were excised with XbaI/BamHI and SacII/XbaI respectively and ligated into the vector pBS-NEO (Figure S2A). The complementation vector was excised with XbaI before transformation of the double deletion mutant strain.

Conidia was inoculated into the 200 mL potato broth to make the initial number of conidia to 10⁵ conidia ml⁻¹ and then cultured at 28°C, 150 rpm for 24 h. Then the mycelium were collected with the nylon membrane, washed two times with 1 M sorbitol and transfer to the 1 M sorbitol containing the 10 mg ml⁻¹ lysing enzyme (Sigma-Aldrich). Then the mycelium was incubated at 28°C, 100 rpm for 3 h to catalyze the cell wall. After that, the protoplast were filtered with nylon membrane and collected by centrifugation with 2,000 rpm at 4°C, washed two times and resuspened with STC buffer (1 M sorbitol, 50 mMTris-Cl, 10 mM CaCl₂, pH 7.4) to the final concentration of 10⁸ ml⁻¹ CFU. For the transformation, 100 μL linearized replacement vector was added into the 200 μL protoplast and the mixture was incubated on ice for 20 min. Then 1 ml 40% PEG dissolved in STC buffer was added into the protoplast mixture and placed for 20 min at 28°C. After the transformation, 5 mL liquid regeneration medium (1 g L⁻¹ yeast extract, 1 g L⁻¹ casein, 6 M sucrose) was added into the protoplast and cultured at 28°C, 100 rpm for 4 h. Then the regenerated protoplast was transfer into the regeneration medium with 1% agar at about 50°C, mixed gently and spread on the petri dish. After the agar concreted, same volume of regeneration medium with 1% agar containing 100 μg ml⁻¹ chlorimuron ethyl or 200 μg ml⁻¹ hygromycin was spread on the upper level to select the transformants. For generation of the double mutants, the Dcl1 deletion mutants were used as the recipient strain to conduct the protoplast preparation and transformation to delete Dcl2. The chlorimuron ethyl-resistant or the hygromycin-resistant strains were isolated and analyzed by PCR with the primer pairs as showed in Figure 1, which are diagnostic for homologous integration of 5′ part and 3′ part. Then the correct transformants were purified by single conidia isolations. Single conidial isolates were obtained by spreading 100 μL of conidial suspension (10⁴ conidia mL⁻¹) on Malt extract agar medium (BD, USA) plates containing 100 μg mL⁻¹ chlorimuron ethyl or 300 μg mL⁻¹ hygromycin. For generation of the complementation mutants, the Dcl1 and Dcl2 double deletion mutants were used as the recipient strain to conduct the transformation. Regeneration medium containing 200 μg ml⁻¹ G418 was used to select the transformants. PCR with the primer pairs 21/22 and 23/24 were used for integration diagnosis. After that, Single conidial isolation was conducted to purify the transformants.

The genomic DNA of the WT and the mutants were extracted and excised with EcoRI. The DNA probes was amplified and labeled by digoxin using digoxigenin-dUTP (Roche) as shown in Figure 1. The DNA band with the hybridized probe was visualized using an enzyme immunoassay and enzyme catalyzed color reaction with NBT/BCIP (Roche).

Wild-type and the mutant strains were grown for 3 day on PDA medium and a disk of hypha with diameter of 1 mm was removed from the growing edge; then the disk of hypha was inoculated on the complete medium (CM) and minimal medium (MM). After culture for 5 days, the diameter was recorded and the growth rate was calculated. Conidia were harvested from C. gloeosporioides strain grown on PDA medium for 12 day, inoculated into 50 mL liquid CM medium to the final concentration of 10³ mL⁻¹, and cultured at 28°C, 150 rpm for desired time. Then the conidia number after incubation for 3 and 4 days were numbered under microscope.

For the pathogenicity assay, conidia of the WT and the mutants were collected, washed two times with ddH₂O and resuspended in a solution of 5% Sabouraud Maltose Broth (Difco) to a final concentration of 2 × 10⁵ conidia mL⁻¹. Droplets (5 μL) of the conidial suspensions were used to inoculate the detached “light green” leaves from rubber tree Varity 73-3-97. The inoculated leaves were kept in a moist chamber at 28°C under natural illumination for 4 days and the disease symptoms were scored. Each treatment contained three replicates of 15 leaves and the entire experiment was repeated three times.

Conidia was prepared as mentioned above and resuspended with sterilized ddH₂O to the final concentration of 2 × 10⁵ conidia mL⁻¹. Onion epidermis was harvested and put on water agar plates. Then the onion epidermis was inoculated with 20 μL of conidia suspension and kept in a moist chamber at 28°C for desired time. After that, the infection structures were analyzed with microscope. The penetration rates of the conidia were quantified after inoculation for 12 h. About 50 conidia were counted under the microscope, and each treatment contained three replication. Sterilized cellophane paper were put on malt extract agar medium, inoculated with 10 μL of conidial suspension, and kept in a moist chamber at 28°C for desired time. The germination were observed by using microscope.

The rubber tree leaves were wounded with sterilized needle. Conidia of WT and mutant strains was prepared as mentioned above, resuspended with 5% Sabouraud Maltose Broth (Difco) to a final concentration of 2 × 10⁵ conidia mL⁻¹. Droplets (5 μL) of the conidial suspensions were used to inoculate the wounded leaves. The inoculated leaves were kept in a moist chamber at 28°C under natural illumination for 24 h. Then the oxidative burst was analyzed by staining the leaves with DAB (Sigma-Aldrich) according to the protocol from Daudi et al. (2012). Wounded leaves inoculated with the only 5% Sabouraud Maltose Broth were used as control check (CK).

Protein extraction, two-dimensional (2D) gel electrophoresis, image analysis and in-gel digestion were carried out according to An et al. (2016). Wild-type and the mutant strains were grown for 3 day on cellophane paper plated on the Malt extract agar medium. The hyphae were collected and disrupted in liquid nitrogen by grinding in a mortar with a pestle. Total protein were extracted with the extraction buffer containing 0.5 M Tris-HCl, pH 8.3, 2% (v/v) NP-40, 20 mM MgCl₂, 2% (v/v) β-mercaptoethanol, and 1 mM PMSF. After removing the cell debris by centrifugation, the supernatant was extracted with an equal volume of Tris-HCl (pH 7.8) buffered phenol. After centrifugation, proteins were precipitated from the final phenol phase with 5 vol of ice-cold saturated ammonium acetate in methanol overnight at −20°C. Then the proteins were collected by centrifugation and washed twice with cold saturated ammonium acetate in methanol and acetone. The precipitate was air-dried for 1 h at 4°C and then solubilized in the thiourea/urea lysis buffer containing 2 M thiourea, 7 M urea, 4% (w/v) CHAPS, 1% (w/v) DTT, and 2% (v/v) carrier ampholytes of pH 3–10. For two-dimensional (2D) gel electrophoresis, aliquots of 650 μg of proteins resolved in 340 μL of sample buffer (7 M urea, 2 M thiourea, 4% (w/v) CHAPS, 1% (w/v) DTT, 2% (v/v) carrier ampholytes (pH 3-10), and 0.001% (w/v) bromphenol blue) were used to rehydrate gel strips (Immobiline DryStrip pH 4–7, 18 cm; GE Healthcare) for 16 h. The first-dimensional IEF, equilibration of the IPG strips and the second-dimensional separation were conducted with the Ettan IPGphor III and the Ettan DALTsix (GE Healthcare) systems according to the manufacturer's instructions. Protein spots whose expression levels changed by >2-fold were excised for protein identification using a MALDI-TOF/TOF mass spectrometer (Ultrafle Xtreme; Bruker Daltonics, Billerica, MA, USA). MS data were uploaded to Mascot for database searching on the Matrix Science (London, UK) public website (http://www.matrixscience.com) and searched against the NCBInr protein database with BIOTOOLS software (v.3.2; Bruker Daltonics). Search parameters were set as taxonomy: Fungi; enzyme: trypsin; max missed cleavages: 1; fixed modifications: carbamidomethyl (C); variable modifications: oxidation (M); peptide mass tolerance: ±100 ppm; fragment mass tolerance: ±0.5 Da. In addition, a peptide charge of 1 + and a monoisotopic mass were selected and the instrument type was set to MALDI-TOF-TOF.

Wild-type and the mutant strains were grown for 3 day on cellophane paper plated on the Malt extract agar medium. The hyphae were collected and disrupted in liquid nitrogen by grinding in a mortar with a pestle, then the RNA was extracted using TRIzol Reagent (Invitrogen). Reverse transcription was conducted with RevertAid RT Reverse Transcription Kit (Thermo Fisher) according to the manufacturer's instructions. Quantitative RT-PCR analysis was performed with the LightCycler 96 System (Roche). The beta-tubulin-1 gene was used as an endogenous control for normalization. Relative expression levels were estimated using the 2^−ΔΔCt method. The primers used for quantitative RT-PCR are listed in Table S1.

Two dicer like protein (DCL) coding gene were identified in C. gloeosporioides using the genome database. The Dcl1 consists of 4,900 bp and codes a protein composed of 1564 amino acids; the Dcl2 consists of 4618 bp and codes a protein composed of 1453 amino acids. To explore their functions in C. gloeosporioides, the nucleotides of two genes were deleted by a replacement strategy as shown in Figure 1. The Chlorimuron ethyl or Hygromycin resistant colonies were analyzed for homologous integration by PCR. To verify the integration locus by PCR, primer pairs with one primer being located outside and one inside the construct were used. As shown in Figure 1, at least three mutants showed the both diagnostic fragments of 5′ and 3′ flanking region. The fragments were sequenced to ensure flawlessness. The results showed that the replacement fragments were correctly integrated into the Dcl loci. The knocking-out mutants were named ΔDcl1 and ΔDcl2 respectively. The results of PCR diagnosis of the double mutant (ΔDcl1ΔDcl2) showed all the diagnostic fragments. All the transformants were purified by single conidia isolation. The detection of WT nuclei of single conidia isolations was implemented by PCR with gene primer pair 17/18 and 19/20. The Southern blot assay indicated that the ΔDcl1, ΔDcl2, and the ΔDcl1ΔDcl2 mutants all showed single homologous integration; and the hybrid bands were consistent with expected DNA sequence lengths (Figures 1C,D). Since the purified single conidia isolates of the three mutants showed the identical phenotypes both in growth rate and pathogenicity, only one strain of each kind of mutant was chosen for detailed studies. Complemented mutant strains were diagnosed by PCR with the primer pairs 21/22 and 23/24. The correct transformants were also purified by single conidia isolation.

The ΔDcl1 and ΔDcl2 showed similar growth rate to WT when cultured on complete medium or minimal medium; while the vegetative growth of ΔDcl1ΔDcl2 was obviously decreased compared with WT (Figure 2A). When cultured in liquid medium, ΔDcl2 generated similar amount of conidia compared with WT, while ΔDcl1 showed an obvious increase in conidiation; for the ΔDcl1ΔDcl2, the conidiation was seriously impaired (Figure 2B). After reintroducing the Dcl1 and Dcl2 nucleotides back into the double deletion mutant, the growth rate of the complemented mutant strain was restored (Figure S2B).

Detached “light green” leaves from rubber tree Varity 73-3-97 were used to determine the pathogenicity of the mutant strains. As shown in Figure 3, about 80% conidia of the WT, ΔDcl1 and ΔDcl2 strains was able to invade the leaves and cause disease, and the lesion diameter were about 6 mm and 9 mm at 2 and 3 dpi. But ΔDcl1ΔDcl2 completely lost the pathogenicity on the rubber leaves. The lesions produced by the complemented mutant strain were nearly similar to those produced by the WT strain (Figure S2C).

To gain a further insight, we tested the conidial germination and penetration on onion epidermis and cellophane paper. After incubation of 12 h, all the conidia of WT germinated and penetrated into the onion epidermis; but the conidia of ΔDcl1ΔDcl2 mutant did not geminated at all; after 24 h, mycelium of WT formed complex networks in onion cells; but the conidia of ΔDcl1ΔDcl2 mutant had just germinated with abnormal germ tubes (Figure 4A). When cultured on the cellophane paper, conidia of ΔDcl1ΔDcl2 mutant germinated at the same rate as WT, although the growth rate was decreased (Figures 4B,C). These results suggested that the penetration ability of ΔDcl1ΔDcl2 mutant was impaired. DAB staining showed that when the wounds were inoculated with the conidia of WT, the rubber leaves could generate significant oxidative burst compared with the CK (inoculated with only Maltose Broth); but that inoculated with ΔDcl1ΔDcl2 mutant did not show significant increase in peroxide (Figure 5).

A comparative proteomic analysis was conducted to identify the proteins that regulated by DCLs. The 2D gel electrophoresis was conducted to separate the proteins isolated from the WT and the ΔDcl1ΔDcl2 mutant; approximately 1680 protein spots were detected on CBB-stained 2D gels. The quantitative image analysis revealed a total of 72 protein spots that showed at least 2-fold down regulation in abundance (P < 0.05) in the double mutant (Figure 6). Besides, only a few protein spots showed up-regulated with abundance change smaller than 2 fold in the mutant. So only the down-regulated protein spots were further analyzed. The 72 down-regulated protein spots were excised and submitted to tandem mass spectrometry and identified by database searching with the Mascot search engine (Table 1). The identified proteins were classified into 9 functional categories based on the FunCatannotation scheme (http://ibis.helmholtz-muenchen.de/funcatDB/), including protein synthesis, cell cycle, chemical metabolism, hydrolytic enzyme, signal transduction, transport, charpones, cell structure and unknown.

To determine whether DCLs regulated the proteins abundance by affecting the transcript levels, the quantitative RT-PCR analysis was conducted. Our results showed that the relative levels of all of the 16 selected genes were significantly decreased; and the results was consistent with that of 2D analysis (Figure 7). These data suggested that the DCLs could directly function by affecting the transcript levels.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.