A Single-Nucleotide Deletion in the Transcription Factor Gene bcsmr1 Causes Sclerotial-Melanogenesis Deficiency in Botrytis cinerea

Botrytis cinerea is an important plant pathogenic fungus with a wide range of host. It usually produces black-colored sclerotia (BS) due to deposition of 1,8-dihydroxynaphthalene melanin in sclerotial melanogenesis. Our previous study (Zhou et al., 2018) reported six B. cinerea isolates producing orange-colored sclerotia (OS) with deficiency in sclerotial melanogenesis. Comparison of ecological fitness (conidia, mycelia, sclerotia), natural distribution, and melanogenesis of selected BS and OS isolates suggests that sclerotia play an important role in the disease cycle caused by B. cinerea. However, the molecular mechanism for formation of the OS B. cinerea remains unknown. This study was done to unravel the molecular mechanism for the sclerotial melanogenesis deficiency in the OS isolates. We found that all the five sclerotial melanogenesis genes (bcpks12, bcygh1, bcbrn1/2, bcscd1) were down-regulated in OS isolates, compared to the genes in the BS isolates. However, the sclerotial melanogenesis-regulatory gene bcsmr1 had similar expression in both types of sclerotia, suggesting the sclerotial melanogenesis deficiency is due to loss-of-function of bcsmr1, rather than lack of expression of bcsmr1. Therefore, we cloned bcsmr1 from OS (bcsmr1OS) and BS (bcsmr1BS) isolates, and found a single-nucleotide deletion in bcsmr1OS. The single-nucleotide deletion caused formation of a premature stop codon in the open reading frame of bcsmr1OS, resulting in production of a 465-aa truncated protein. The transcription activity of the truncated protein was greatly reduced, compared to that of the 935-aa full-length protein encoded by bcsmr1BS in the BS isolates. The function of bcsmr1OS was partially complemented by bcsmr1BS. This study not only elucidated the molecular mechanism for formation of orange-colored sclerotia by the spontaneous mutant XN-1 of B. cinerea, but also confirmed the regulatory function of bcsmr1 in sclerotial melanogenesis of B. cinerea.


INTRODUCTION
Botrytis cinerea is an important plant pathogenic fungus in cool and temperate regions. It can infect more than 1,400 plant species, including about 500 commercially important vegetables, fruits, and ornamentals, causing gray mold disease (Elad et al., 2016). B. cinerea usually produces gray-colored conidia and hyphae, and black-colored sclerotia on the surface of the infected plant tissues, as it can synthesize melanin and deposit that dark pigment in the cell wall of hyphae, conidia, and sclerotia (Zeun and Buchenauer, 1985;Doss et al., 2003;Zhang et al., 2015;Schumacher, 2016).
Melanins are multifunctional compounds (Cordero and Casadevall, 2017). It is well-recognized that melanin deposition in the cell wall can improve fungal survival through enhancing fungal adaptation to environmental stresses, such as UV irradiation, desiccation, enzymatic lysis, toxic chemicals, antagonists, and predators (Bell and Wheeler, 1986;Butler and Day, 1998). In some plant pathogenic fungi, such as Colletotrichum legenarium and Pyricularia oryzae, melanin is a virulence factor, as it can facilitate appressorium-mediated penetration into plant tissues (Kubo et al., 1982;Howard and Valent, 1996;Chen et al., 2004). In other plant pathogenic fungi, such as Alternaria brassicicola, however, melanin biosynthesis was found to be negatively associated with virulence (Cho et al., 2012). In B. cinerea, melanin is dispensable for pathogenicity, and instead, it may affect longevity of conidia and sclerotia as proposed in previous studies Schumacher, 2016).
Sclerotia are generally regarded as the dormancy structure for survival under adverse environmental conditions. Sclerotia produced by B. cinerea can germinate to produce conidia. However, importance of the sclerotia as the primary infection source remains controversial in the life cycle of B. cinerea, as they are not frequently observed in infected tissues of many diseased plants (Williamson et al., 2007). Zhou et al. (2018) reported six field isolates of B. cinerea producing orange-colored sclerotia (OS). The sclerotia of the OS isolate XN-1 were found to be deficient in sclerotial melanogenesis, compared to the blackcolored sclerotia produced by isolate B05.10 of B. cinerea (Zhou et al., 2018). The orange-colored sclerotia were greatly reduced for survival, compared to the black-colored sclerotia. Meanwhile, Zhou et al. (2018) also found that the OS B. cinerea rarely occurred in nature. Therefore, DHN melanin deposition on the sclerotia of B. cinerea plays an important role in ecological fitness of sclerotia. However, the molecular mechanism responsible for formation of the orange-colored sclerotia by the field isolates of B. cinerea remains unknown. This study was conducted to address this question. Here, we report that a single-nucleotide deletion in bcsmr1 caused formation of a premature stop codon, resulting in formation of a truncated protein with loss-offunction in regulation of sclerotial melanogenesis.

Fungal Isolates and Culture Media
A total of eight isolates of B. cinerea were used in this study ( Table 1). Among these isolates, XN-1, S417, and T417 are OS isolates, whereas B05.10, HS016, S59, WXto2-2, and XN087 are BS isolates. All these fungal isolates were maintained in 20% glycerol (v/v) at −80 • C for the long-term storage. Working cultures of each isolate were prepared by transferring the mycelia of that isolate in storage on potato dextrose agar (PDA) and the resulting cultures were incubated in the dark at 20 • C for 5 days. Seven cultural media (MYA, PDA, PDB, PRM, YPDA, SD/-Trp, and SD/-Trp-His) were used in this study. The composition of these media was listed in Table S1.

Gene Cloning and Sequence Analysis
Six B. cinerea isolates (XN-1, S417, T417, HS016, WXt02-2, XN087) were separately incubated on cellophane film (CF)-PDA at 20 • C for 3 days (Zhou et al., 2018). The mycelial mats were collected from the cultures, and used for extraction of genomic DNA (gDNA) using the CTAB method (Zeng et al., 2014) and the total RNA using the Trizol R reagents (TaKaRa Biotechnol. Co. Ltd., Dalian, China). The gDNA was used as template to PCR-amplify the DNA sequences of bcsmr1 as well as bcpks12/13, bcbrn1/2, and bcscd1 with the specific primers and thermal programs (Tables S2, S3). The PCR product was separated by agarose gel electrophoresis. The target DNA band in the gel was purified and inserted into the pMD18-T vector (TaKaRa

Prokaryotic Expression of bcsmr1
The full-length cDNA of bcsmr1 was cloned by reverse transcription (RT)-PCR from the OS isolate XN-1 and the BS isolate B05.10 using the primer pair bcsmr1-etF/bcsmr1-etR ( Table S2). The primers contain a Not I-restriction site. The resulting cDNA was separated by agarose gel electrophoresis and purified from the gel. Then, the cDNA was digested with Not I to release the target fragments of bcsmr1 from XN-1 (designated as bcsmr1 OS ) or B05.10 (designated as bcsmr1 BS ). The two cDNA fragments were separately inserted into the plasmid pET28a(+) (Beijing TransGen Biotechnol. Co. Ltd., Beijing China) at the Not I site and two new plasmids, namely pET28a-bcsmr1 OS and pET28a-bcsmr1 BS (both had a 6× His tag), were then regenerated ( Figure S1). The two plasmids were separately transformed into the competent cells of E. coli BL21 (DE3) pLysS (Promega, Madison, WI, USA). Positive clones for the two plasmids were screened based on growth on LB containing kanamycin (50 µg/mL). They were shake-incubated and induced for production of the proteins BcSMR1 OS and BcSMR1 BS using the procedures described by Lou et al. (2015). The bacterial proteins were separated by 8% SDS-PAGE and visualized by staining with Coomassie brilliant blue G-250.

Western Blotting
The proteins in E. coli transformed with pET28a-bcsmr1 OS or pET28a-bcsmr1 BS were prepared using the procedures described by Lou et al. (2015) and were separated by 8% SDS-PAGE. After electrophoresis, the proteins in the gel were transferred by electrophoresis blotting (20 V, 30 min) to a piece of Bio-Rad R polyvinylidene fluoride (PVDF) membrane (6 × 4 cm, length × width) in Trans-Blot R SD Semi-Dry Electrophoretic Transfer Cell. The membrane was washed in TBST buffer (20 mmol/L Tris-base, 150 mmol/L NaCl, 0.05% Tween 20), followed by soaking for 2 h in 5% (w/v) Difco TM skim milk solution alone and then in 5% skim milk solution amended with 0.2% (w/v) anti-His tag antibody (Sangon Biotechnol. Co. Ltd., Shanghai, China) for another 2 h. After that, the membrane was washed again in the TBST buffer for three times, 10 min each time, followed by soaking for 2 h in 5% skim milk solution amended with 0.1% (w/v) alkaline phosphatase-conjugated goat antimouse IgG (Sangon Biotechnol. Co. Ltd., Shanghai, China). After another washing in the TBST buffer, the membrane was soaked in 10 mL Western Blue R Stablized Substrate for Alkaline Phosphatase (Promega) to visualize the specific protein bands on the membrane. Finally, the membrane was dipped in water to stop the reaction and photographed to show the hybridzation.

Gene Transcription Activation Assay
The transcription activity of bcsmr1 OS and bcsmr1 BS was determined using Matchmaker TM GAL4 two-hybrid system 3 (Clontech Laboratories, 2007). The cDNA sequences for bcsmr1 BS in the BS isolate B05.10 (2,080 bp long), bcsmr1 OS in the OS isolate XN-1 (2,079 bp long), and bcactA (coding for actin) in B05.10 (1,128 bp long) were separately amplified by RT-PCR with the total RNA extracts from two isolates as templates using the specific RT-PCR primer sets and the specific thermal programs (Tables S2, S3). Meanwhile, the DNA sequence containing the GAL4 activation domain (AD GAL4 , 408 bp long) was PCR amplified from the plasmid pGADT7 (TaKaRa) with the primer set GAL4-F/GAL4-R ( Table S2).
The resulting cDNA fragments for bcsmr1 OS , bcsmr1 BS , and bcactA, and the DNA fragment of AD GAL4 were purified from a 1% agarose gel after electrophoresis. They were separately inserted into the plasmid pGBKT7 containing the DNA binding domain of GAL4 (BD GAL4 ) using ClonExpress R II One Step Cloning Kit (Vazyme Biotechnol. Co. Ltd., Nanjing, China). The resulting plasmids pBD GAL4 -AD GAL4 (positive control), pGBKT7 (negative control 1), pBD GAL4 -BcACTA (negative control 2), pBD GAL4 -BcSMR1 OS , and pBD GAL4 -BcSMR1 BS were separately transformed into the yeast cells of strain AH109 using the LiAc/SS carrier DNA/PEG transformation method (Clontech Laboratories, 2007). Selection of the yeast mutants harboring the plasmids and detection of their transcription activity were done using the procedures described by Lou et al. (2015).

Determination of the Transcripts of the Melanogenesis-Related Genes
The transcripts of bcsmr1, bcygh1, bcpks12/13, bcbrn1/2, and bcscd1 in the sclerotia of the B. cinerea isolates were determined using qRT-PCR with the specific primers and thermal cycles (Tables S2, S3). All the investigated isolates were incubated on PDA at 20 • C. The sclerotial primordia, immature sclerotia, and mature sclerotia were collected from the 6-, 8-, and 15day-old cultures of each isolate, respectively. The total RNA in the sclerotial samples was extracted using TaKaRa RNAiso Reagent Kit (TaKaRa) and was subjected to treatment with DNase I (TaKaRa) to eliminate DNA contamination in the extracts. The treated RNA was then used as templates in qRT-PCR to determine the transcripts of bcactA (reference), bcsmr1, bcpks12/13, bcygh1, bcbrn1/2, and bcscd1. The transcript levels of each gene were calculated by the Ct method (Zeng et al., 2014). For normalization of the data, the transcript level of each gene in the sclerotial primordia of isolate B05.10 was considered as 1.0. The scale was used to calibrate the transcript level of that gene in the immature and mature sclerotia of B05.10, as well as in the sclerotial primordia, immature sclerotia, and mature sclerotia of other isolates. The qRT-PCR assay was independently repeated three times.

Complementation of the Sclerotial Melanogenesis in XN-1 with bcsmr1 bs
The complementation plasmid pKS1004-BS was constructed based on the plasmid pKS1004 containing the hygromycinresistance gene (hph) (Zeng et al., 2014). The full-length cDNA of bcsmr1 BS from B05.10 was ligated at the downstream of the Aspergillus nidulans promoter Polic by fusion PCR. The resulting DNA fragment was inserted into the plasmid pKS1004 using the protocol and the reagents in ClonExpress R II One Step Cloning Kit (Vazyme Biotechnol. Co. Ltd., Nanjing, China). The plasmid was then transformed into the protoplasts of XN-1 (Zeng et al., 2014). The transformed protoplasts were regenerated at 20 • C on the PRM medium (Table S1) amended with 50 µg/mL hygromycin B. The emerging colonies (isolates) were individually transferred to PDA. Meanwhile, isolates B05.10 and XN-1 were inoculated on PDA as two controls. The cultures (isolates) were incubated at 20 • C for 35 days for production of sclerotia. Six complementation mutants were selected based on formation of brown-to black-colored sclerotia (indicating sclerotial melanogenesis). Integration of bcsmr1 BS in the genomes of the complementation mutants was confirmed by PCR using the primer set Olic-F/bcsmr1R ( Table S2). XN-1 was used as reference in the specific PCR assay. Two representative complementation mutants, namely XN-BS11 and XN-BS14, were selected for determination of the copy number of bcsmr1 by Southern blotting also with XN-1 as a reference. The bcsmr1-specific DNA probe was generated by PCR from the genomic DNA of B05.10 with the primer set bcsmr1-jcF/bcsmr1R (Table S2). In Southern blotting, the gDNA of XN-1, XN-BS11, and XN-BS14 was extracted and digested with EcoRI. The resulting DNA fragments were separated by electrophoresis and detected by the DNA probe, which was labeled with the reagents in DIG High Prime DNA Labeling and Detection Starter Kit II (Roche R , Mannheim, Germany).

Data Analysis
The procedure UNIVARIATE in the SAS software (SAS Institute, Cary, NC, USA, v. 8.0, 1999) was used to analyze the data on relative expression levels of each gene. Average values of each gene between the OS (3 isolates) and BS (4 isolates) groups, and between each complementation mutant and XN-1/B05.10 were compared using Student's t-test at P < 0.05 or P < 0.01.

Expression of the Genes for Sclerotial Melanogenesis Is Suppressed in the Orange-Colored Sclerotia, Compared to that in the Black-Colored Sclerotia
To analyze the melanin-biosynthesis ability in orange-colored sclerotia of B. cinerea, expression of six melanogenic enzymescoding genes, including bcpks12/13, bcygh1, bcbrn1/2, and bcscd1, in the sclerotia of seven isolates of B. cinerea (Table 1) was determined by qRT-PCR using the specific primers and thermal programs (Tables S2, S3). The results showed that all the seven investigated isolates (OS isolates: XN-1, S417, T417; BS isolates: B05.10, HS016, XN087, WXt02-2) had low expression levels of bcpks13 at all sclerotial developmental stages (Figure 1). Thus, FIGURE 1 | Relative expression values of six melanogenic enzyme-coding genes (bcpks12/13, bcygh1, bcbrn1/2, and bcscd1) in the sclerotia of B. cinerea isolates. The sclerotial primordia, immature sclerotia, and mature sclerotia of each isolate were collected from 6-, 8-, and 15-day-old PDA cultures (20 • C), respectively. The results are the representative of three independent tests. Bars indicate standard deviations of the means. In each graph, "*" at P < 0.05 and "**" at P < 0.01 between the OS and BS isolates (as two groups) according to Student's t-test.
bcpks13 did not seem to be involved in the formation of the orange-colored sclerotia.
However, significant differences were found between the OS and BS isolates in expression levels of the polyketide synthase gene bcpks12 (Figure 1). Although all the seven investigated isolates showed the increased expression pattern in maturation of sclerotia, the expression levels were significantly lower in the OS isolates than in the BS isolates. In the sclerotial primordia (white-colored) from the 6-day-old PDA cultures (20 • C), the expression level of bcpks12 was low with the relative expression values (REVs) ranging from 0.3 to 1.4 for the OS isolates and from 1.0 to 22.6 for the BS isolates. In the immature sclerotia (gray-colored) from the 8-day-old PDA cultures and in the mature sclerotia (black-and orange-colored in the BS and OS isolates, respectively) from 15-day-old PDA cultures, the bcpks12 expression levels slightly increased by 2.6 and 7.8-folds, respectively, in the OS isolates, whereas were largely increased by 40.4 and 121.8-folds, respectively, in the BS isolates (Figure 1).
For the other sclerotial melanogenesis genes (bcygh1, bcbrn1/2, bcscd1), the OS isolates differed from the BS isolates both in expression pattern and in expression level (Figure 1). While the BS isolates showed an increased expression pattern, the OS isolates showed a decreased expression pattern. In the sclerotial primordia of the OS isolates, the REVs ranged from 0.01 to 0.03 for bcygh1, 1.5 to 2.1 for bcbrn1, 1.3 to 2.6 for bcbrn2, and 0.4 to 0.8 for bcscd1. In the sclerotial primordia of the BS isolates, the REVs ranged from 0.1 to 1.0 for bcygh1, 1.0 to 3.4 for bcbrn1, 1.0 to 6.9 for bcbrn2, and 0.3 to 1.1 for bcscd1. In the immature sclerotia and mature sclerotia, expression of these genes was decreased by 43.2-88.5% in the OS isolates, whereas it was increased by 8 to 90-folds in the BS isolates, compared the corresponding REVs of these genes in the sclerotial primordia of B05.10.
To rule out the possibility that suppressed expression of bcpks12, bcygh1, bcbrn1/2, and bcscd1 in the sclerotia of the OS isolates is caused by mutations of these genes, the DNA sequences of these genes were cloned from the BS and OS isolates mentioned above (Table S4). Alignment analysis showed that the identity of the DNA sequences of each gene and the amino acid sequences of each protein in the seven above-mentioned isolates of B. cinerea was higher than 98% (Table S5). Single-amino acid polymorphisms were detected in the amino acid sequences of each protein ( Figure S2). However, none of the polymorphic amino acids was found to be associated with formation of the orange-colored sclerotia. Therefore, suppressed expression of bcpks12, bcygh1, bcbrn1/2, and bcscd1 in the sclerotia of the OS isolates of B. cinerea is probably not caused by the point mutations in these genes. It might be caused by some regulatory genes for the sclerotial melanogenesis.

The OS Isolates Have a Single-Nucleotide Deletion in bcsmr1
To confirm whether the abnormal expression pattern of the melanin biosynthesis genes was caused by some regulatory malfunction, the transcription factor (TF) gene bcsmr1 related to the melanin biosynthesis was cloned from three OS isolates (XN-1, S417, T417) and four BS isolates (HS016, S59, WXt02-2, XN087) by PCR using the specific primers and specific thermal cycles (Tables S2, S3). The DNA sequences for bcsmr1 in these isolates were submitted to GenBank and were assigned with the accession numbers from KU743104 to KU743109, and KX098785 (Table S4). They were aligned with the DNA sequence of bcsmr1 in the reference isolate B05.10 (Accession Number: Bcin02g08760 in the website http://fungi.ensembl.org/Botrytis_ cinerea). The results showed that bcsmr1 in the OS isolates (designated as bcsmr1 OS ) differed from bcsmr1 in other five BS isolates (designated as bcsmr1 BS ) in length of the open reading frame (ORF), 3,096 bp long for bcsmr1 BS , whereas 3,095 bp long for bcsmr1 OS (Figure 2A). The single-nucleotide deletion in bcsmr1 OS is located at the position of nt 1,524, where guanine (G) is present in bcsmr1 BS , but absent in bcsmr1 OS . Thus, the singlenucleotide deletion was designated as G1524X. In spite of the single-nucleotide deletion, the identity of the DNA sequences in the ORF region of bcsmr1 OS and bcsmr1 BS in the seven isolates was higher than 98% ( Table S5).
The single-nucleotide deletion in bcsmr1 OS caused change of the triplet codon at the position nt 1,522 to 1,524 from ATG (for methionine or M) in bcsmr1 BS to ATA (for isoleucine or I) in bcsmr1 OS . The subsequent triplet codon at the position nt 1,525 to 1,527 was also changed from ATA (for isoleucine or I) in bcsmr1 BS to TAG (stop) in bcsmr1 OS (Figure 2A). As a consequence, the ORF of bcsmr1 OS was deduced to encode a protein with 465 amino acid (aa) residues (designated as BcSMR1 OS ). It is a truncated protein of the 935aa protein encoded by the ORF of bcsmr1 BS (designated as BcSMR1 BS ). Alignment analysis showed that both BcSMR1 465 and BcSMR1 935 have the DNA binding domain at the Nterminus, consisting of two Cys 2 His 2 -zinc finger motifs and a Zn 2 Cys 6 -binuclear motif ( Figure 2B). BcSMR1 OS has a small portion (32 aa long) of the fungal specific transcription factor domain at the C-terminus without the nuclear localization signaling, whereas BcSMR1 BS has the full fungal specific TF domain (298 aa long) with the nuclear localization signaling region. Alignment analysis also showed that there are six point mutations (T125I, A174D, V181I, L223S, S232P, M465I) in BcSMR1 BS and BcSMR1 OS ( Figure 2B) and five of these mutations (T125I, A174D, V181I, L223S, S232P) were found in both BcSMR1 BS and BcSMR1 OS . Therefore, they are probably not associated with formation of the orange-colored sclerotia. The other mutation, namely M465I, was detected only in BcSMR1 OS , but it was not detected in BcSMR1 BS . This mutation is caused by the single-nucleotide-deletion-mediated onenucleotide shift (Figure 2A). Therefore, the single-nucleotide deletion is probably important for formation of the orangecolored sclerotia by the OS isolates of B. cinerea. In spite of the point mutations, the identity of bcsmr1 OS and bcsmr1 BS is higher than 99% (Table S5).
Prokaryotic expression of bcsmr1 OS and bcsmr1 BS in E. coli was done to confirm formation of BcSMR1 OS through the singlenucleotide deletion-mediated truncation of BcSMR1 BS . The fulllength cDNA of bcsmr1 OS in XN-1 and bcsmr1 BS in B05.10 were separately cloned by RT-PCR using the specific primers and specific thermal program (Tables S2, S3). The resulting cDNA fragments were separately introduced into the expression vector pET28a(+) (Figure S1). The resulting two plasmids, namely pET28a-bcsmr1 OS and pET28a-bcsmr1 BS , were then constructed and separately transformed into E. coli BL21 (DE3) pLysS. The transformed bacterial clones were selected, followed by shake-incubation in LB medium for production of BcSMR1 OS and BcSMR1 BS under induction by isopropyl β-D-1-thiogalactopyranoside (IPTG). The bacterial proteins extracted from different E. coli cultures were detected by 8% SDS-PAGE. Results showed that two differentially-expressed proteins were consistently detected in the positive transformants of E. coli (data not show). The molecular weight was ∼66 and 120 kD (both had a 6× His tag) in the transformants with pET28a-bcsmr1 OS and pET28a-bcsmr1 BS , respectively. The 54-kD difference in molecular weight between the two proteins is close to the 53.2-kD difference in molecular weight between the putative BcSMR1 OS and BcSMR1 BS , which had the predicted molecular weight of 51.3 and 104.5 kD, respectively. Western blotting analysis using the anti-His tag antibody further confirmed production of the 66and 120-kD proteins by bcsmr1 OS and bcsmr1 BS , respectively, in E. coli (Figure 3).
The Transcription Activity of bcsmr1 OS Is Greatly Suppressed, Compared to that of bcsmr1 Bs Both the OS isolates and the BS isolates had an increased expression pattern of bcsmr1 in sclerotia (Figure 4). We hypothesized that compared to bcsmr1 BS , bcsmr1 OS might be suppressed for activating transcription of the downstream genes for sclerotial melanogenesis. To test this hypothesis, the transcription activity of bcsmr1 OS and bcsmr1 BS was determined in Saccharomyces cerevisiae AH109 using Matchmaker TM GAL4 Two Hybrid System 3 (Clontech Laboratories, 2007). Five plasmids (pBD GAL4 , pBD GAL4 -BcACTA, pBD GAL4 -AD GAL4, pBD GAL4 -BcSMR1 OS , and pBD GAL4 -BcSMR1 BS ) were constructed. Four plasmids (pBD GAL4 -BcACTA, pBD GAL4 -AD GAL4, pBD GAL4 -BcSMR1 OS , and pBD GAL4 -BcSMR1 BS ) contain two domains, namely the DNA-binding domain (e.g., BD GAL4 ) and the activation domain (e.g., AD GAL4 , BcACTA, BcSMR1 OS , BcSMR1 BS ). The remaining plasmid pBD GAL4 contains BD GAL4 alone. The plasmids were separately transformed into the yeast cells and the resulting transformants were screened on two auxotrophic media SD/-Trp and Frontiers in Microbiology | www.frontiersin.org FIGURE 3 | Western blotting detection of the E. coli proteins encoded by bcsmr1 BS and bcsmr1 OS using the anti-His tag antibody. The treatments "bcsmr1-BS"and "bcsmr1-OS" represent the proteins encoded by bcsmr1 BS , and bcsmr1 OS , respectively. The treatment "pET28a" represents the protein encoded by the blank plasmid pET28a alone without bcsmr1 BS or bcsmr1 OS . M = PageRuler TM prestained protein ladder. SD/Trp-His. On SD/-Trp, all the mutants showed growth. On SD/-Trp-His, however, the mutants differed in growth ( Figure 5). While the transformants harboring either pBD GAL4 or pBD GAL4 -BcACTA failed to grow on SD/-Trp-His, the transformants of pBD GAL4 -AD GAL4 , pBD GAL4 -BcSMR1 OS , and pBD GAL4 -BcSMR1 BS grew on this medium and produced blue color on the colonies when X-α-gal was added. This result indicates that AD GAL4 , bcsmr1 BS , and bcsmr1 OS can activate transcription of the MEL1 gene coding for α-galactosidase. Relatively, the transformants of pBD GAL4 -AD GAL4 and pBD GAL4 -BcSMR1 BS grew better than the transformant harboring pBD GAL4 -BcSMR1 OS . The blue color on the colonies of the transformants of pBD GAL4 -AD GAL4 and pBD GAL4 -BcSMR1 BS was darker than that on the colonies of the transformant of pBD GAL4 -BcSMR1 OS . Therefore, compared to bcsmr1 BS , bcsmr1 OS is greatly suppressed for transcription of MEL1.

Complementation of bcsmr1 OS with bcsmr1 BS Restored Sclerotial Melanogenesis in XN-1
To confirm responsibility of the single-nucleotide deletion in bcsmr1 OS for sclerotial melanogenesis deficiency, the full-length FIGURE 4 | Relative expression values of the transcription factor gene bcsmr1 in the sclerotia of B. cinerea. The sclerotial primordia, immature sclerotia, and mature sclerotia of each isolate were collected from 6-, 8-, and 15-day-old PDA cultures (20 • C), respectively. The results are the representative of the three independent tests. Bars indicate standard deviations of the means. "*" at P < 0.05 and "**" at P < 0.01 between the OS and BS isolates (as two groups) according to Student's t-test.
cDNA of bcsmr1 BS from B05.10 was used to complement the function of bcsmr1 OS in XN-1. A complementation plasmid pKS1004-BS containing the cDNA of bcsmr1 BS under the control of the constitutive A. nidulans PoliC promoter was successfully constructed on the basis of the plasmid pKS1004 harboring the hygromycin-resistance gene hph (Zeng et al., 2014). The plasmid was introduced into XN-1 by protoplastmediated transformation. Ten transformants were obtained based on hygromycin resistance. After incubation at 20 • C on PDA for 35 days, XN-1 formed orange-colored sclerotia, as it has defect in sclerotial melanogenesis, whereas B05.10 formed black-colored sclerotia, as it has normal sclerotial melanogenesis (Figure 6). Four of these transformants formed orange-colored sclerotia like those formed by XN-1, indicating that the sclerotial melanogenesis in these transformants was not restored at all. In contrast, six other transformants (XN-BS1, XN-BS2, XN-BS11, XN-BS12, XN-BS14, XN-BS16) formed partiallymelanized sclerotia, which were characterized by formation of color mosaic on the sclerotial surface (Figure 6). These six transformants could be classified into two groups based on the color mosaic on the sclerotia. Three transformants (XN-BS11, XN-BS12, XN-BS16) showed a brown-gray mosaic pattern, whereas the other three transformants (XN-BS1, XN-BS2, XN-BS14) showed a gray-black mosaic pattern. This result indicates that the sclerotial melanogenesis in these transformants was at least partially restored, compared to the complete sclerotial melanogenesis in B05.10 and the defective sclerotial melanogenesis in XN-1. Meanwhile, sporogenic germination (indicating lack of dormancy) was observed on the orange sclerotia produced by XN-1 and on the brown-grayish sclerotia produced by the transformants XN-BS11, XN-BS12, and XN-BS16 (Figure 6). The sporulation appeared to be vigorous on the sclerotia of XN-1, whereas was sparse on the sclerotia of the three transformants. In contrast, the black-colored sclerotia produced by B05.10 and the gray-colored sclerotia produced by the transformants XN-BS1, XN-BS2, and XN-BS14 did not germinate at all. This result indicates that the dormancy deficiency of the orange sclerotia (germination-active) of XN-1 due to lack of melanin was partially restored in the partiallymelanized sclerotia produced by the complementation mutants.
The six complementation transformants were confirmed for integration of bcsmr1 BS in their genomes by PCR using the primer set Olic-F/bcsmr1R (Table S2) (data not shown). Two of these transformants (XN-BS11, XN-BS14) were further determined for bcsmr1 BS integration in their genomes by Southern blotting using the bcsmr1-specific DNA probe ( Figure S3A). XN-1 was used as the reference isolate in Southern blotting. The result showed that the three isolates had one to three hybridization bands, one band in XN-1, two bands in XN-BS11, and three bands in XN-BS14 ( Figure S3B). This result indicates that the genomes of XN-1, XN-BS11, and XN-BS14 harbor one (bcsmr1 OS ), two (bcsmr1 OS + bcsmr1 BS ), and three (bcsmr1 OS + bcsmr1 BS + bcsmr1 BS ) copies of bcsmr1, respectively.
Expression of the six melanogenic genes (bcpks12/13, bcygh1, bcbrn1/2, bcscd1) in the sclerotia of isolates XN-1, XN-BS11, and XN-BS14 was determined by qRT-PCR. The two transformants were similar to XN-1 both in expression pattern and in expression level of bcpks13, which is not required for sclerotial melanogenesis according to Schumacher (2016). All the three isolates showed a decreased expression pattern and low expression levels of bcpks13 in sclerotial maturation.
In contrast, the two transformants differed greatly from XN-1 either in expression pattern or in expression level of the other five sclerotial melanogenesis genes (bcpks12, bcygh1, bcbrn1/2, bcscd1). For bcpks12 and bcygh1, the two transformants were similar to XN-1 in expression pattern, but differed from XN-1 in expression level (Figure 7). The two genes showed an increased expression pattern in maturation of the sclerotia produced by the three isolates. In the sclerotial primordia, the REVs ranged from 1.0 to 49.8 for bcpks12 and from 1.0 to 10.0 for bcygh1. In the immature sclerotia and mature sclerotia, the REVs of the two genes were largely increased by 12 to 350-folds in the two transformants, whereas were slightly increased by 3 to 12-folds in XN-1.
For bcbrn1, the two transformants differed from XN-1 both in expression pattern and in expression level (Figure 7). In XN-1, the REV was reduced from 1.0 in the sclerotial primordia to < 0.3 in the immature sclerotia/mature sclerotia. In XN-BS11 and XN-BS14, however, the REVs were increased in maturation of the sclerotia. XN-BS11 had the REV increase from 0.6 in the sclerotial primordia to 2.6 in the immature sclerotia, and 7.4 in mature sclerotia. XN-BS14 had a REV increase from 2.9 in the sclerotial primordia to 10.8 in the immature sclerotia, and 13.1 in the mature sclerotia.
For bcbrn2 and bcscd1, the two transformants were also similar to XN-1 in expression pattern, but differed from XN-1 in expression level (Figure 7). All the three isolates showed a decreased expression pattern in maturation of the sclerotia. In the sclerotial primordia, the three isolates had the REVs of 0.9-2.3 and 0.9-1.8 for bcbrn2 and bcscd1, respectively. In the immature sclerotia and mature sclerotia, while XN-1 had negligible expression of both genes, XN-BS11 had the REVs of 0.1-0.6 for bcbrn2 and 0.4-1.7 for bcscd1. In general, XN-BS14 had higher REVs of bcpks12, bcygh1, bcbrn1/2, and bcscd1 than the corresponding REVs for XN-BS11 (Figure 7). Meanwhile, the expression pattern of bcpks12, bcygh1, and bcbrn1/2 in the mutants XN-BS11 and XN-BS14 resembled that in the BS isolate B05.10, although the expression levels of these genes at each stage were significantly (P < 0.05 or P < 0.01) lower than those of B05.10 (Figure 7). This result further confirmed that the singlenucleotide deletion in bcsmr1 is the major reason for formation of the orange-colored sclerotia. Zhou et al. (2018) first reported natural occurrence of orangecolored sclerotial (OS) isolates of B. cinerea infecting strawberry and tomato in central China. The OS isolates are natural mutants or variants of B. cinerea, compared to the normal isolates with FIGURE 7 | Relative expression values of the six melanogenic enzyme-coding genes (bcpks12/13, bcygh1, bcbrn1/2, bcscd1) in the sclerotia of isolate XN-1 of Botrytis cinerea, two complemented mutants of XN-1 (XN-BS11, XN-BS14) and B05.10. The sclerotial primordia, immature sclerotia, and mature sclerotia of each isolate were collected from 6-, 8-, and 15-day-old PDA cultures (20 • C), respectively. The results are the representative of three independent tests. Bars indicate standard deviations of the means. *P < 0.05 and **P < 0.01 (Student's t-test) between each mutant and XN-1; P < 0.05 and P < 0.01 (Student's t-test) between each mutant and B05.10.

DISCUSSION
formation of black-colored sclerotia. Microscopic observation showed that both orange-and black-colored sclerotia of B. cinerea are quite similar in structure, consisting of the cortex layer (outside) with compact parenchymatous cells and the medulla (inside) with loose prosenchymatous cells. However, the sclerotia of the two types differ greatly in melanin deposition in the wall of the epidermal cells. While the BS isolates have normal melanin deposition in the black-colored sclerotia, the OS isolates are deficient in melanin deposition in the orange-colored sclerotia (Zhou et al., 2018). Our previous study also showed that the OS isolates occur rarely in nature, as the sclerotia produced by the OS isolates of B. cinerea survived poorly, compared to the sclerotia produced by the BS isolates (Zhou et al., 2018). Therefore, mutation of the sclerotial color from black to orange greatly changed the ecological fitness of B. cinerea. Zeun and Buchenauer (1985) reported that B. cinerea produces DHN melanin, as tricyclazole, an inhibitor for DHN melanin biosynthesis, can delay sclerotial melanogenesis in B. cinerea. Schumacher (2016) elucidated the molecular pathway for biosynthesis of DHN melanin in B. cinerea. The result showed that bcpks12 is required for sclerotial melanogenesis, bcpks13 is required for conidial melanogenesis, and bcbrn1/2 and bcscd1 are required for both sclerotial melanogenesis and conidial melanogenesis (Schumacher, 2016). The present study revealed that the OS isolates differed greatly from the BS isolates either in expression pattern or in expression level of bcpks12, bcygh1, bcbrn1/2, and bcscd1. The expression levels of these genes in the immature and mature sclerotia of the three OS isolates were significantly lower than those of the four BS isolates (Figure 1). This molecular evidence corroborates our previous study in microscopic observation that the orange-colored sclerotia was due to lack of melanin deposition (Zhou et al., 2018).
Bcsmr1 and other CMR1 orthologs are similar in structure. They contain a DNA-binding domain with two Cys 2 His 2 zinc finger motifs and one Zn(II) 2 Cys 6 binuclear cluster motif at the N-terminus, and a fungal specific transcription factor (TF) domain and a nuclear localization signaling (NSL) region at the C-terminus ( Figure 2B). Tsuji et al. (2000) found that the DNA-binding domain in CMR1 is necessary for melanization in C. lagenarium. Fetzner et al. (2014) reported that CmrA encoded by CmrA can bind to the promoter region of pksA (coding for polyketide synthase), thereby initiating expression of pksA and the subsequent melanogenesis in A. alternata.
This study demonstrated that both BS and OS isolates of B. cinerea had an increased expression pattern of bcsmr1 in the sclerotia (Figure 4). The result is consistent with the finding of Schumacher (2016), who found that expression of bcsmr1 is associated with sclerotial development of B. cinerea. It implies that failure to regulate the sclerotial melanogenesis by bcsmr1 OS in the OS isolates of B. cinerea is not due to expression of this gene. Instead, it may be caused by loss-of-function of bcsmr1 OS .
In order to elucidate the mechanism for the loss-of-function of bcsmr1 OS , we cloned bcsmr1 OS and bcsmr1 BS from selected isolates in this study, and compared the DNA sequences of bcsmr1 OS with those of bcsmr1 BS . Three OS isolates (XN-1, S417, T419) have a single-nucleotide deletion in bcsmr1 OS (Figure 2B). Occurrence of single-nucleotide deletions has been frequently reported in animals (Persuy et al., 1999;Yang et al., 2003) and plants (Takano et al., 2014;Lou et al., 2015), whereas it has been rarely reported in fungi, including B. cinerea (Pihet et al., 2009).
Previous studies readily indicated that B. cinerea can mutate in response to extreme environments such as presence of UV irradiation and fungicides (Leroux et al., 2002;Kretschmer et al., 2009). These mutagens usually cause formation of singlenucleotide polymorphisms (SNPs) in populations of B. cinerea (Kretschmer et al., 2009;Schumacher et al., 2012Schumacher et al., , 2013. Most SNPs in certain genes such as mbc1 coding for β-tubulin have been found to be closely associated with sensitivity or resistance to fungicides (Leroux et al., 2002;Kretschmer et al., 2009). A few SNPs in other genes such as bcvel1 coding for the VELVET protein cause formation of a premature stop codon, thereby resulting in production of truncated proteins and change of some important phenotypes such as virulence and sclerotial formation (Schumacher et al., 2012(Schumacher et al., , 2013. Whether or not these mutagens can cause formation of the single-nucleotide deletion in bcsmr1 remains unknown. Taking together all the information about the low occurrence frequency of the OS isolates of B. cinerea in nature and the reduced ecological fitness of the OS isolates, we can draw a conclusion that the single-nucleotide deletion observed in this study is important for B. cinerea. Understanding of the mutagens for formation of that single-nucleotide deletion in bcsmr1 will be helpful for designing a novel strategy for control of B. cinerea. This study demonstrated that the single-nucleotide deletion in bcsmr1 OS caused formation of a premature stop codon, which results in production of a truncated 465-aa protein BcSMR1 OS (Figures 2, 3). In contrast, the full-length bcsmr1 BS in the BS isolates codes for a 935-aa protein BcSMR1 BS (Figures 2, 3). BcSMR1 OS contains the complete DNA-binding domain, but has only the partial transcription factor domain without the nuclear signaling localization region. In S. cerevisiae AH109, bcsmr1 OS had reduced activity in activating transcription of the MEL1 gene, compared to bcsmr1 BS (Figure 5). This result suggests that bcsmr1 OS cannot be fully functional in activating transcription of the sclerotial melanogenesis genes in B. cinerea. Like other transcription factors, bcsmr1 is transcribed in the nuclei and the resulting transcripts are translocated into the cytoplasm, where they are translated to the protein BcSMR1. Without the nuclear signaling localization domain, BcSMR1 OS may not be so easy as BcSMR1 BS in re-entrance into the nuclei from the cytoplasm. Without the full transcription factor domain, BcSMR1 OS may not be so fully functional as BcSMR1 BS in activating transcription of the downstream genes for sclerotial melanogenesis even it enters into the nuclei somehow. As a consequence, transcription of the sclerotial melanogenesis genes regulated by BcSMR1 OS is reduced or blocked, compared to BcSMR1 BS . Thus, the sclerotial melanin biosynthesis in the OS isolates of B. cinerea is suppressed, thereby causing formation of the orange-colored sclerotia.
Introduction of the full-length bcsmr1 BS from the BS isolate B05.10 partially complemented the loss-of-function of bcsmr1 OS in the OS isolate XN-1. Six complementation mutants formed partially-melanized sclerotia (Figure 6). The mutants XN-BS11 and XN-BS14 showed enhanced expression of bcpks12, bcygh1, bcbrn1/2, and bcscd1 in the immature and mature sclerotia in comparison to their progenitor XN-1 (Figure 7), confirming the previous finding by Schumacher (2016) that bcsmr1 has the regulatory function for the sclerotial melanogenesis in B. cinerea. Ectopic complementation with the full-length bcsmr1 BS increased expression of bcpks12, bcygh1, and bcbrn1 in sclerotia and the expression pattern was similar to that found in the BS isolate B05.10. However, this study also found that the complementation had negligible effects on expression of bcbrn1/2 and bcscd1 (Figure 7). These results suggest that bcsmr1 has different regulatory effects on expression of these genes. Coincidently, bcbrn2 and bcscd1 along with bcpks13 form one gene cluster (in chromosome No. 3) located at a different chromosomal position from the other sclerotial melanogenesis genes. Bcsmr1, bcpks12, and bcygh1 form a gene cluster located at chromosome No. 2, and bcbrn1 is located at chromosome No. 4 (Schumacher, 2016). Difference in chromosomal position for bcsmr1, bcbrn2, and bcscd1 may be responsible for the low regulatory efficiency of bcsmr1 for bcbrn2 and bcscd1. Further studies are necessary to determine the activity of bcsmr1 BS in activating transcription of different sclerotial melanogenesis genes.
This study used the full-length bcsmr1 from the BS isolate B05.10, instead of the deleted nucleotide, to restore the phenotype of formation of black-colored sclerotia by the OS isolate XN-1. This complementation strategy may mask the possible effects of the point mutations of bcsmr1 on formation of orange-colored sclerotia. A few point mutations were identified in bcsmr1 among the eight isolates of B. cinerea (5 BS isolates and 3 OS isolates) (Figure 2). However, none of the point mutations was found to be associated with formation of the orange-colored sclerotia. Therefore, the result from the complementation experiment in this study at least strengthened importance of the single-nucleotide deletion in loss-of-function of bcsmr1 for sclerotial melanogenesis. Future studies are necessary to complement bcsmr1 OS with the deleted nucleotide.
Schumacher (2016) reported that bcygh1 involves in conidial melanogenesis of B. cinerea, as BcYGH1 encoded by bcygh1 can convert the product catalyzed by BcPKS13 encoded by bcpks13, which is required only for conidial melanogenesis. Fan et al. (2017) reported that Vayg1, an ortholog of bcygh1, is required for microsclerotia formation and melanogenesis of microsclerotia in Verticillium dahlia, the causal agent of verticillium wilt of cotton. In this study, we found that like the essential genes for sclerotial melanogenesis (bcpks12, bcbrn1/2, bcscd1), bcygh1 was greatly suppressed for expression in the orange sclerotia of XN-1, compared to expression of bcygh1 in the black sclerotia of B05.10 (Figure 1). However, when bcsmr1 BS was introduced into XN-1, bcygh1 expression was greatly increased (Figure 7). This result implies that besides Bcltf1/2 (two transcription factors), bcsmr1 may regulate expression of bcygh1. It also suggests that the expression of bcygh1 is closely associated with sclerotial melanogenesis in B. cinerea. Further studies are necessary to clarify the interaction between bcsmr1 and bcygh1 and to elucidate how the interaction affects the sclerotial melanogenesis in B. cinerea.

AUTHOR CONTRIBUTIONS
YZ and JZ: designed research; LY and MW: provided new agents and analyzed the data; YZ, WC, and GL: wrote the paper.

ACKNOWLEDGMENTS
This study was financially supported by the National Natural Science Foundation of China (Grant No. 31571955) and China's Special Fund for Agro-Scientific Research in the Public Interest (Grant No. 201303025). We appreciate the kind help from Prof. Zhonghua Ma (Zhejiang University, Hangzhou, China) for providing isolate B05.10 of B. cinerea.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb. 2017.02492/full#supplementary-material Figure S1 | A schematic diagram showing the procedure for construction of the plasmids used for prokaryotic expression of bcsmr1 OS and bcsmr1 BS from Botrytis cinerea isolates XN-1 and B05.10, respectively.  Figure S8 prepared by Schumacher et al. (2014); (B) Five schematic diagrams showing differential amino acids in the five melanogenic enzymes among the investigated isolates of B. cinerea. Two polyketide synthases: BcPKS12 and BcPKS13 encoded by bcpks12 and bcpks13, respectively; two reductases: BcBRN1 and BcBRN2 encoded by bcbrn1 and bcbrn2, respectively; and one dehydratase: BcSCD1 encoded by bcscd1. Abbreviations for the conserved domains: KS, β-ketoacyl synthase domain; AT, acyl transferase domain; ACP, acyl carrier protein domain; TE, thioesterase domain; ENR, enoyl-ACP reductase doamin; SD, scytalone dehydratase domain. The GenBank accession numbers for the polypeptides-coding genes were listed in Table S4.