Reactive Oxygen Species Play a Role in the Infection of the Necrotrophic Fungi, Rhizoctonia solani in Wheat

Rhizoctonia solani is a nectrotrophic fungal pathogen that causes billions of dollars of damage to agriculture worldwide and infects a broad host range including wheat, rice, potato and legumes. In this study we identify wheat genes that are differentially expressed in response to the R. solani isolate, AG8, using microarray technology. A significant number of wheat genes identified in this screen were involved in reactive oxygen species (ROS) production and redox regulation. Levels of ROS species were increased in wheat root tissue following R. solani infection as determined by Nitro Blue Tetrazolium (NBT), 3,3'-diaminobenzidine (DAB) and titanium sulphate measurements. Pathogen/ROS related genes from R. solani were also tested for expression patterns upon wheat infection. TmpL, a R. solani gene homologous to a gene associated with ROS regulation in Alternaria brassicicola, and OAH, a R. solani gene homologous to oxaloacetate acetylhydrolase which has been shown to produce oxalic acid in Sclerotinia sclerotiorum, were highly induced in R. solani when infecting wheat. We speculate that the interplay between the wheat and R. solani ROS generating proteins may be important for determining the outcome of the wheat/R. solani interaction.


Introduction
The soil-borne pathogen, Rhizoctonia solani is a damaging necrotrophic fungus causing billions of dollars of losses to agriculture worldwide [1]. This pathogen has a broad host range and infects major crops including barley, canola, legumes, potato, rice and wheat. However, crop losses due to soil-borne pathogens are often underestimated and their economic importance is expected to continue to rise due to the increased implementation of reduced or no-tillage farming practices. R. solani is classified into fourteen reproductively incompatible anastomosis groups (AGs). One of these, AG8, is a devastating pathogen causing bare patch of cereals, brassicas and legumes [2,3].
interaction. In this study we have investigated the interaction between wheat and R. solani AG8 and studied responses from both the plant and pathogen side. We found that a significant number of wheat genes identified in this screen were involved in ROS production and redox regulation. Biochemical staining also confirmed that ROS species were induced during R. solani/wheat infection. In addition, homologous genes to known ROS/redox related genes were identified from R. solani and, in some cases, their expression was shown to be affected during plant infection.

Maintenance of R. solani AG8
The R. solani isolate AG8 (WAC10335) was obtained from the Department of Agriculture and Food (www.agric.wa.gov.au). R. solani AG8 was grown on fresh potato-dextrose agar plates (PDA, Sigma-Aldrich, St Louis, MO) at room temperature for 1 week. Millet seeds were used for R. solani infection and storage. Moist millet seeds were autoclaved and inoculated with R. solani grown in PDA, at 22°C in the dark for 2 weeks, with occasional mixing. For longer-term storage, inoculated millet seeds were dried overnight in a sterile laminar flow hood and stored at -80°C.
Wheat growth and R. solani infection Two wheat cultivars have been used in this study; Wyalkatchem is a standard wheat cultivar commonly grown in Western Australia [34], and Chinese Spring is the reference cultivar used for the genome sequencing project [35]. Triticum aestivum cv Chinese Spring was sourced from Dr. Nicolas Taylor, University of Western Australia, Perth, Australia, and T. aestivum cv Wyalkatchem was sourced from CSIRO, Black Mountain, Canberra. Plants were grown in a growth cabinet at a temperature range of 21°C to 24°C, with a 16 hour light photoperiod (100 μM.m -1 . sec-1 ) and were treated with R. solani AG8 or a mock treatment as described below.
Fine vermiculite was infected by placing four R. solani infected millet seeds on top of each 5cm x 5cm pot which was covered with foil, contained in a plastic bag for 1 week and left to grow at room temperature. Unsterilised wheat seedlings were germinated on damp blotting paper for one day (Zadoks growth stage GS05) and transferred to pre-colonised pots [36]. After 2, 3 and 4 days, root and shoot tissue from Chinese Spring wheat were collected (3 biological replicates for  each tissue/time point; each replicate was taken from a separate individual plant) For the pathogenicity studies, two-week-old Wyalkatchem and Chinese Spring seedlings were removed from vermiculite and rinsed in water to remove debris and the shoots were weighed for five plants from each cultivar. R. solani pathogenicity was scored by weighing the shoot mass on both Wyalkatchem and Chinese Spring seedlings with mock and R. solani treatment, 2 weeks post-infection. Five biological replicates were used for each treatment and Student's t-test was performed on this data.

Infection conditions for analysis of R. solani gene expression
Sterile nitrocellulose membrane (Immobilon 82mm, poresize 0.45μm, Millipore, Millipore. com) was placed on the surface of a PDA plate and 1 cm 2 plug of inoculum from a prior PDA culture of fungus was placed on top of the membrane. The fungus was allowed to grow at 24°C for 1 week until it had almost covered the surface of the 8.5cm diameter nitrocellulose circle. Wheat seeds were surface sterilised in sodium hypochlorite solution (1.7% available chlorine) with shaking for 15 minutes and rinsed 3 times in sterile water. The seeds were added to a petri dish containing filter paper moistened with sterile water and transferred to 4°C for three days, then transferred to 24°C in the dark for 3 days to germinate (Zadoks growth stage GS07). The seedlings were removed from the Petri dishes and added to fresh plates containing 15 mL of minimal medium. The nitrocellulose membrane and attached mycelium was removed from the PDA plate and placed mycelium side down onto the wheat seedlings in minimal medium. The plates were incubated at an average temperature of 24°C for 4 days or 7 days with 16 hours of light (100 μM.m -1 .sec-1 ) prior to harvesting. The membrane and attached mycelium was removed from the plates, leaving behind the wheat seedlings. The mycelium was peeled from the membrane, blotted dry and frozen in liquid nitrogen. Vegetative fungal samples were treated as above with the omission of wheat seedlings and sampled after 7 days.
One microgram of total RNA from 4-day-old R. solani infected or mock treated wheat seedlings were sent to AGRF (Australian Genomic Research Facility, www.agrf.org.au) to hybridize using the Agilent Wheat Gene Expression Microarray (www.agilent.com) which contained glass slides formatted with four high-definition 44K probes (60mer oligonucleotides). Three biological replicates of each treatment was included. The content was pre-2008 release; sourced from RefSeq31, Unigene 53, TIGR 11 and TIGR TA5. This Agilent Wheat Oligo microarray with 37,826 probes is registered as GPL13636 in the Gene Expression Omnibus (GEO) at the National Center for Biotechnology Information (NCBI). All expression tag contig sequences used for the design of probes have been published online (KOMUGI, http://www.shigen.nig.ac. jp/wheat/komugi/array/probe/download.jsp).

Bioinformatic analysis of microarray data
Functional annotations were derived from mapping the probes of the Agilent Wheat Microarray to the latest wheat transcriptome sequences (IWGSC MIPS v2.2) [42,43]. Probe sequences were mapped to the transcriptome by BLAT (-fastMap) [44] and functional annotations were assigned to probes based on IWGSC MIPS v2.2 InterPro, Pfam and GO functional annotations assigned to transcriptome sequences. Results of this mapping is presented in S1 Dataset.

ROS detection
Unsterilised one-day-old Wyalkatchem wheat seedlings (Zadoks GS05) germinated on wet filter paper were added to R. solani AG8 infected vermiculite or mock vermiculite and then grown for a further 3 days. Seedlings were removed from vermiculite and debris was rinsed off the plant. Whole seedlings were treated with 2mM Nitro Blue Tetrazolium (NBT, Sigma-Aldrich) in 20mM phosphate buffer, pH6.1 to detect O 2 .-. The seedlings were then rinsed in water to remove excess NBT and photographed using a digital camera (Nikon, D3200). For H 2 O 2 detection by 3,3'-diaminobenzidine (DAB) staining, seeds of Chinese Spring and Wyalkatchem wheat were sterilized in 5% hypochlorite, washed 3x in sterile water and placed on sterile filter paper to germinate. Five-day-old seedlings were then transferred to vermiculite that had been pre-treated with R. solani AG8-infested millet seeds or a mock treatment for seven days at 24°C until harvest. DAB staining was performed using non-acidified 3,3'-diaminobenzidine (Sigma-Aldrich) at a concentration of 1mg/mL, prepared according to the method of Daudi and O'Brien, [45]. AG8-infected and mock-infected wheat roots were gently washed to remove vermiculite and then stained for 30 minutes in DAB solution. Roots were washed in water to remove excess DAB before imaging using a digital camera (Nikon, D3200) oror using a dissecting microscope (Nikon, SMZ-10). Dissecting microscope images were taken at 30x magnification using AmScope software (www.amscope.com).
Biochemical H 2 O 2 measurement was performed using the titanium sulphate (Sigma-Aldrich) treatment and O 2 .was measured by NBT. Measurement of NBT reduction, a method used for the determination of O 2 .was described by Doke [46] and the measurement of H 2 O 2 by titanium sulphate was originally reported by Snell and Snell [47]. Approximately 100 mg of root tissue, not including 1 cm of root tips, were snap frozen in liquid nitrogen, and ground before suspending in 500 μL of 100 mM phosphate buffer, pH 6.8. The suspension was spun on a

Identification of wheat genes differentially expressed following R. solani infection
A growth cabinet based pathosystem between wheat and R. solani was modeled on work previously developed for R. solani/Arabidopsis interactions [49]. By day 4, the pathogenic effect of R. solani infection on wheat plants could be visualised, with stunted root growth (Fig 1).
RNA extracted from Chinese Spring wheat root tissue from day 4 post-infection with R. solani AG8 grown on minimal media was used to identify differentially expressed genes compared to mock treatment using DNA microarray analyses. The arrays were obtained from Agilent Technologies, and have 44K, 60 mer probes which have been used to study various aspects of wheat biology [50,51]. Bioinformatic analysis using a !2 fold change cutoff and adjusted p value 0.05 identified 83 genes that were induced and one gene that was repressed. A complete set of microarray data from this study was deposited with the GEO repository under the accession number GSE74671, and the differentially expressed genes are listed in S1 Dataset.
The only gene that was identified as being repressed (5-fold and P adjusted rate <0.05), was a gene homologous to phospholipase D that has been shown to play a role in regulating lipid metabolism and signalling, which can impact on hormone signalling pathways that mediate plant defence responses and abiotic stresses (reviewed by [52]). Of the 83 genes that were found to be induced (> 2 fold and P adjusted rate <0.05), a majority fell into the ROS/redox regulation area or were linked to detoxification, for example five induced genes displayed homology to germin, a gene that catalyzes the conversion of oxalate in the presence of O 2 into H 2 O 2 while a number of PR genes were also found to be induced (Table 1, S1 Dataset).
To confirm the expression of genes determined by the microarray experiment, four highly induced genes were tested for RNA expression over 2-4 days post-infection with R. solani AG8 in root and shoot tissue using RT-qPCR (Fig 2). Analysis of root tissue at day 4 following AG8 treatment showed high expression of germin (GLP1), UDP-glucose/galactose 4 epimerase (UGE1) and PR10 verifying the Agilent array data results and they were predominantly expressed in the root compared to the shoot. While TaGlu (Beta 1,3-glucanase), was strongly induced at Day 3, there was no significant induction at Day 4 in roots. The TaGlu gene sequence was derived from the draft wheat genome sequence but did not produce a perfect match, suggesting other family members in wheat may be contributing to the microarray results.
The interaction between R. solani and wheat resulted in the production of ROS As a number of ROS-related genes were identified from the array, it was of interest to determine the ROS status of the wheat plants during R. solani infection. Infected seedlings and controls were stained with DAB to measure H 2 O 2 or NBT to measure O 2 .-. In both cases more ROS were seen in the AG8 treated seedling compared to mock (Fig 3A, 3B

Expression of ROS related genes in R. solani
Given the response in wheat to R. solani AG8 infection involved both the induction of a number of ROS-related genes and increases in ROS, it was of interest to analyse ROS related responses in R. solani. ROS related genes were identified in R. solani by nucleotide/protein blast searches from previously characterised fungal genes involved in ROS. Seven R. solani ROS-related homologous genes were identified from the coding regions of the R. solani genome [3]; noxA, noxB, noxC, noxD, TmpL, SOD and OAH. Fig 4 shows the relative expression of each R. solani gene following infection in wheat using RT-qPCR. The four nox homologous genes did not show any change in expression and the SOD-homologous gene was repressed whilst those homologous to TmpL and OAH showed significant induction compared to R. solani AG8 growing vegetatively (Fig 4).

Discussion
R. solani is a pathogenic fungus that causes physical changes including stunted root growth of the host, transcriptional changes and ROS production. Using differential expression analysis Analysis of selected wheat genes in response to R. solani AG8 using RT-qPCR. One day old wheat seedlings were infected with R. solani AG8 for 2, 3 or 4 days, and RNA from root and shoot tissue were tested for relative RNA expression for A) GLP1 (Germin), B) UGE1, C) PR10 and D) TaGlu (glucan endo-1,3-beta-glucosidase GII precursor). RNA expression was normalised to the ADP-ribosylation factor R wheat gene. Stars indicates significant differences for mock vs R. solani infection (log transformed Student's t-test, P <0.05). doi:10.1371/journal.pone.0152548.g002 between non-infected and R. solani infected wheat, 83 genes were identified that were induced and 1 gene that was repressed. Reduced number of differentially expressed genes were detected using the R. solani AG8 and Chinese Spring wheat Agilent arrays compared to other systems such as the Arabidopsis/R. solani AG8 (resistant-7 days post infection) Affymetrix analysis, where 501 genes were identified that were induced and 205 genes were identified that were repressed using similar stringency criteria [15]. In addition, the Arabidopsis/R. solani AG2-1 (susceptible-7 days post infection) Affymetrix analysis identified 357 genes that were induced and 229 genes that were repressed [15]. It remains to be determined if there are fewer wheat than Arabidopsis genes affected by R. solani infection or the wheat probes used in the Agilent array do not accurately reflect the full scope of all wheat genes. It was of interest to note that the genes that were tested for RT-qPCR expression were found to be induced at a higher rate than that seen in the microarrays, suggesting that the microarray may have limitations with regards to sensitivity. Alternatively, because wheat has a hexaploid genome, mRNA expression from other homologs may be masking gene expression changes and therefore RNA sequencing may be required to uncover homolog specific differences in gene expression in response to R. solani.
Twenty wheat genes were identified from the wheat microarray that were induced greater than ten-fold, and sixteen of these genes show homology to genes with known functions or which have been previously associated with specific conditions such as biotic stress. These included genes homologous to the pathogen-related (PR) genes such as TaPR5 (thaumatinlike), TaPR10 and TaGlu (glucan endo-1,3-beta-glucosidase GII precursor). The PR5 and PR10 genes have been previously shown to be induced during stripe rust infection [53,54] although the induction was significantly less than what was observed with R. solani infection. The second most abundant gene was identified as a homolog to UDP-glucose/galactose 4 epimerase (UGE), which may have a role in energy production (glycolysis) or inositol production for sequestering phosphate away from bacteria/fungi. Multiple genes from the glucosyltransferase family were identified as being induced during the R. solani infection. Glucosyltransferase genes have been associated with detoxifying mycotoxins in other fungi Ornithine decarboxylase is required for Parastagonospora nodorum virulence in wheat [55], and the fungal pathogen Fusarium graminearum may exploit the induction of ornithine as a cue for mycotoxin production during wheat infection [56]. Agmatine coumaroyltraserase has been shown to be the first step of biosynthesis of an antifungal derivative in barley [57]. Ent-copalyl diphosphate synthase functions to synthesise gibberellin/phytoalexin [58] and laccase is an oxidase that has a role in the regulation of lignin production [59].
Two of the genes induced over 10 fold were germin-like proteins (GLP). The plant responds to the toxic effects of oxalic acid by producing the enzyme germin, which catalyzes the conversion of oxalate in the presence of O 2 into H 2 O 2. Proteins with 30 to 70% amino acid identities with characterised germin protein sequences are designated as germin-like proteins (GLPs) [60]. The H 2 O 2 that is released in this reaction may play a role in plant defense as antimicrobial agents, and cross-linking of cells walls or through acting as a messenger in signal transduction pathways linked to oxidative stress signalling (reviewed by [61]). It was of interest to observe that the R. solani OAH gene was also induced in response to R. solani/wheat infection. The S. sclerotiorum-P. vulgaris interaction also resulted in an induction of an OAH homolog [62]. The R. solani OAH gene encodes a gene homologous to the enzyme oxaloacetate acetylhydrolase which has been shown to produce oxalic acid [63]. Oxalic acid is believed to act in pathogenesis through acidification of host tissues and sequestration of calcium from host cell walls and has a role in both suppressing and inducing the oxidative burst of the host plant [64][65][66]. Isolates of R. solani are known to produce oxalic acid and the virulence of rice infecting strains was correlated with increased oxalic acid production [67]. In addition, over-expression of a germin like oxalate oxidase in rice lead to increased resistance against R. solani [68]. In our study both wheat GLP and R. solani OAH were induced at both time points used, with no significant difference in expression level between time points. This co-induction of pathogen and host genes with contrasting functions suggests the interplay between wheat GLP and R. solani OAH may be important in the interaction of R. solani with wheat.
Wheat and barley germins have been shown to be induced by abiotic stress and biotic stress, although the amount of induction does not necessary correlate with the level of biotic resistance observed [69][70][71]. In a number of instances, overexpression of germin in Arabidopsis, canola, soybean and tomato can lead to enhanced resistance to fungal pathogens [72][73][74][75][76], however reverse genetic studies of oxalate oxidase in transgenic wheats did not show a role of oxalate oxidase in response to bacterial blight [77].
The enigma of ROS in plant-pathogen interactions, is that pathogens often require ROS for pathogenicity and developing infection structures, and the host requires ROS for resistance. ROS is used by the host and pathogen both directly and indirectly, as for example ROS toxicity is used to kill plant cells or fungal hyphae, while ROS also serves as an important signal molecule. Also, both pathogen and host have detoxifying enzymes to reduce the ROS damage within the organism. The temporal and spatial expression of when, where and by whom ROS is produced and its function during the wheat/R. solani interaction may hold the key to help understand the respective role(s) ROS plays for the plant and pathogen. Therefore, future research directions are likely to focus on developing experiments to improve resolution and quantification of ROS to address the specific timing and location of production during pathogen infection of plants and to be able to unequivocally distinguish ROS produced by the plant from that produced by the pathogen.