This is the accepted version of the article: Aragona, Maria; Campos-Soriano, Lidia; Piombo, Edoardo; [et al.]. «Imaging the invasion of rice roots by the bakanae agent Fusarium fujikuroi using a GFP-tagged isolate». European Journal of Plant Pathology, (May 2021). DOI 10.1007/s10658-021-02301-z This version is available at https://ddd.uab.cat/record/241549 under the terms of the license Imaging the invasion of rice roots by the bakanae agent Fusarium fujikuroi using a GFP-tagged isolate Maria Aragona1, Lidia Campos-Soriano2, Edoardo Piombo3,4, Elena Romano5, Blanca San Segundo2,6, Davide Spadaro3,4, Alessandro Infantino1 1 Council for Agricultural Research and Economics (CREA), Research Centre for Plant Protection and Certification (CREA-DC), Via C.G. Bertero 22, 00156 Rome, Italy. 2 Centre for Research in Agricultural Genomics (CRAG) CSIC-IRTA-UAB-UB, Edifici CRAG, Campus UAB Bellaterra (Cerdanyola del Vallés), 08193 Barcelona, Spain. 3 Dept. Agricultural, Forest and Food Sciences (DISAFA) and AGROINNOVA - Centre of Competence, University of Torino, Largo Braccini 2, 10095 Grugliasco (TO), Italy 4 AGROINNOVA – Centre of Competence for the Innovation in the Agro-environmental Sector, University of Turin, Largo Braccini 2, 10095 Grugliasco, TO, Italy. 5 Centre of Advanced Microscopy “P. Albertano”, Department of Biology, University of Rome “Tor Vergata”, 00133 Rome, Italy. 6 Consejo Superior de Investigaciones Científicas, Barcelona, Spain Corresponding author: Maria Aragona, maria.aragona@crea.gov.it ID ORCID of the authors: Maria Aragona: 0000-0002-1320-2141 Lidia Campos-Soriano: 0000-0002-0814-7934 Edoardo Piombo: 0000-0003-2830-1967 Elena Romano: 0000-0002-8501-2590 1 Blanca San Segundo: 0000-0001-7409-3172 Davide Spadaro: 0000-0001-5207-9345 Alessandro Infantino: 0000-0003-0048-1257 Concise title: Rice roots infection by gfp-tagged Fusarium fujikuroi Acknowledgements This work was supported by the AGER Foundation (grant 2010‐2369), project RISINNOVA ‘Integrated genetic and genomic approaches for new Italian rice breeding strategies’; by the Ministerio de Ciencia, Innovación y Universidades/FEDER–Agencia Estatal de Investigación (RTI2018-101275-B-I00), the CERCA Programme (“Generalitat de Catalunya”), and MINECO (“Severo Ochoa Programme for Centres of Excellence in R&D” 2016-2019, SEV‐2015‐0533). Author contributions Maria Aragona designed the project, performed the root infection and analysis and wrote the manuscript; Lidia Campos-Soriano performed the genetic transformation of F. fujikuroi; Edoardo Piombo performed the expression analyses and wrote the manuscript; Elena Romano performed LSCM analysis; Alessandro Infantino contributed to design the research work and cared the mycological part; Davide Spadaro and Blanca San Segundo contributed to design the research work and writing the manuscript. All authors commented on previous versions of the manuscript, read and approved the final manuscript. 2 ABSTRACT Fusarium fujikuroi (teleomorph Gibberella fujikuroi) is the main seed-borne pathogen of rice, the causal agent of bakanae, a disease that in the last years has become of increasing economical concern in many Italian rice growing areas. A virulent F. fujikuroi isolate was tagged with the green fluorescent protein (gfp) gene using Agrobacterium tumefaciens mediated transformation, and the virulence of the GFP isolate has been confirmed. Little is known about the early interaction of the pathogen with its host, in this work fungal development during the F. fujikuroi/root interaction was analysed by Laser Scanning Confocal Microscopy (LSCM), by using the GFP isolate obtained. The infection of rice roots was investigated from 48 h to 8 days post-inoculation both in resistant and susceptible cultivars. Roots of resistant genotype seem to trigger a hypersensitive response at the infection site and LSCM analysis of root sections allowed the visualization of fungal growth within host tissues. Fungal growth occurred both in the resistant and the susceptible cultivar, even if it was less abundant in the resistant one. Expression analysis of Chitinase1, a gene involved in fungal pathogenesis, was investigated by qPCR on the F. fujikuroi infected rice roots. Chitinase1 expression increased greatly upon infection in the resistant cultivar Selenio. Keywords: genetic transformation, Agrobacterium tumefaciens, confocal laser scanning microscopic analysis, gene expression INTRODUCTION Bakanae is a rice disease caused by the hemibiotrophic fungal pathogen Fusarium fujikuroi. It was originally observed in Japan in 1928 (Ito and Kimura 1931), but it is now present in several countries in America, Europe, Asia, Oceania and Africa (Amatulli et al. 2010; Carter et al. 2008; Chen et al. 2016; Desjardins et al. 2000; Jeon et al. 2013; Karov et al. 2005; Khan et al. 2000; Kim et al. 2015; Zainudin et al. 2008). The meaning of Bakanae is “foolish seedling”, and it is due to the main symptom of the disease: the elongation and thinning of internodes, inducing frail stems and abnormal 3 height, thin leaves, and grains entirely or partially empty. The altered plant morphology is due to the production of gibberellic acids (GAs) by F. fujikuroi, the only Fusarium species capable of GAs biosynthesis (Ou 1985). GAs are not essential for fungal growth and development but, controlling jasmonic acid-mediated plant immune responses, they probably contribute to the virulence of F. fujikuroi (Wiemann et al. 2013; Siciliano et al. 2015). Although F. fujikuroi is predominantly a seed borne pathogen, it also survives in soil and diseased plant debris (Ou 1985). Seeds can become infested when conidia, produced on diseased plants, use the wind and water splash to reach neighbouring panicles at flowering. In a recent work Sunani et al. (2019), studying the infectious structures, penetration and colonization of F. fujikuroi in seeds and seedlings of rice, showed that infection through floret is the dominant pathway to seed infection. The localization of the pathogen could be both inside and on the outside of the seed, being predominant in the lemma and palea, followed by embryo (Kumar et al. 2015). Seeds can also be contaminated by the fungus at harvest, when they can be reached by conidia produced on diseased and dead plants. A third source of seedborne infection is represented by spores and mycelium contaminating the water used to stimulate germination in soaked seeds (Karov et al. 2009). Both ascospores and conidia can also infect seedlings through the roots and crown, colonising both the intracellular and intercellular spaces of the rice root: the fungus invades the plant without producing visible symptoms, so that F. fujikuroi can be found in apparently healthy seeds (Kanjanasoon 1965). Bakanae disease is increasing in the main rice-producing areas worldwide (Infantino et al. 2017; Amatulli et al. 2010). Losses in rice production caused by bakanae depend on climate, rice cultivars and pathogen strain, ranging from 3% to 15% in Thailand (Kanjanasoon 1965), 2% - 20% in Macedonia (Karov et al. 2005), 20% - 50% in Japan (Ito and Kimura 1931), up to 52% in Bangladesh (Hossain et al. 2013), and up to 75% in Iran (Saremi et al. 2008). The most common Bakanae management is based on thermal seed treatment and the use of fungicides, but F. fujikuroi resistance to various fungicides has been reported (Chen et al. 2016). The need of developing new control measures is therefore increasing. The identification of new sources of resistance to F. fujikuroi was 4 based on the screening of large collections of rice germplasm and allowed to map several quantitative trait loci (QTLs) on rice chromosomes (Chen et al. 2019; Volante et al. 2017). The development of simple sequence repeat (SSR) markers and mating type analysis allowed to detect F. fujikuroi genetic variability at population level (Valente et al. 2016), which is of fundamental importance for screening of resistance. In this work, we focused on the analysis of early stages of root infection by a fluorescent F. fujikuroi isolate, with the aim to unravel the differences between the susceptible and the resistant rice cultivars facing pathogen infection and colonization. Interaction between pathogens and host plants have been extensively studied using fluorescent reporter proteins. Organisms that express genes encoding fluorescent reporter proteins are frequently used to monitor pathogen behaviours in plant tissues under various physiological conditions (Lagopodi et al. 2002; Oren et al. 2003). The advantage of GFP as a reporter is that it allows in vivo imaging of fungal hyphae during its interaction with the host plant. Hyphae of gfp-expressing fungal strains can be visualized in living tissue in real time, using fluorescence microscopy without extensive manipulation. Compared to many fungal pathogens, such as Aspergillus spp. and other Fusarium species, the lack of efficient technologies for genetic manipulation has become a major obstacle for the development of F. fujikuroi molecular research (Cen et al. 2020). However, a polyethylene glycol (PEG)-mediated transformation of protoplasts has been used to introduce the gfp and the red fluorescent protein (rfp) gene into F. fujikuroi for visualizing interaction with biocontrol agents (Watanabe et al. 2007; Kato et al. 2012) and the early root colonization of a GA-producing wild-type and a GA-deficient mutant strain (Wiemann et al. 2013). Recently, a gfp-expressing F. fujikuroi isolate, obtained by PEG transformation, has been used to analyse rice infection of susceptible and resistant cultivars at the basal stem level, by confocal microscopy analysis (Lee et al. 2018). No study so far, an A. tumefaciens-based method has been developed for transformation of F. fujikuroi. By this way we transformed four virulent F. fujikuroi strains by using a gfp-expressing vector, and one of the GFP-tagged isolates obtained was used to visualize and analyse the infection and colonization processes at root level in susceptible and resistant rice cultivars, by confocal 5 microscopy. Quantification of expression in the rootlets of chitinase 1, a gene related to the response to bakanae disease, was also performed. MATERIALS AND METHODS Fungal strains and growth conditions Four virulent F. fujikuroi strains were selected inside a collection of more than 300 isolates stocked at CREA-DC and previously used for a study of population structure analysis (Valente et al. 2016). The Ff 192, Ff 297, Ff 364 and Ff 1550 isolates were chosen among the most virulent ones and belonging to different haplotypes. After transformation by the gfp vector, as described below, four isolates named Ff 192-GFP, Ff 297-GFP, Ff 364-GFP and Ff 1550-GFP were obtained, and they are all listed in Online Resource 1. Fusarium fujikuroi isolates were grown on potato dextrose agar (PDA) or potato dextrose broth (PDB) at 23˚C, in the case of transformed isolates hygromycin (Hyg) at concentration of 100 µg ml-1 was added to the media. Generation of F. fujikuroi strains expressing the gfp gene The four selected F. fujikuroi isolates were transformed with the plasmid pCAMgfp (kindly provided by A. Sesma, John Innes Center, UK) (Sesma and Osbourn 2004). The pCAMgfp plasmid contains the sgfp gene (Chiu et al. 1996) under the control of the ToxA promoter from Pyrenophora tritici- repentis (Lorang et al. 2001) and the hygromycin phosphotransferase (hph) gene as the selectable marker gene. The pCAMgfp plasmid was introduced into the A. tumefaciens AGL-1 strain, the virulent strain required for fungal transformation. F. fujikuroi transformation was carried out using the A. tumefaciens AGL-1-transformed strain following the protocol previously described (Campos- Soriano and San Segundo 2009; Campos-Soriano et al. 2013) with minor modifications. Co- cultivation was performed at 25˚C and selection was done at 28˚C. PDA medium plus hygromycin B (250 µg ml-1 final concentration) was used as selective medium to grow the F. fujikuroi transformed isolates. Fungal colonies were transferred to 24-well plates containing the selective medium to test 6 the effective transformation. A stereomicroscope (Olympus SZX16) with 480-nm excitation and 500 to 550-nm emission filter block was used to verify GFP-transformed fungal colonies. The stability of transgene integration and gfp expression of transformants were tested by sub-culturing them for five generations on PDA medium and then transferring them again on selective PDA medium containing 100 µg ml-1 hygromycin B. The number of pCAMgfp copies integrated into the genome of transformants has been assessed by qReal Time-PCR, using the primers Hyg588F and Hyg588R, listed in Online Resource 2. The PCR mix was composed of 10 µl of SensiMix 2x (Bioline), 2 µl of primer mix (forward and reverse, 5 µM of each primer) and 4 µl of nuclease free water. To each sample 2 µl of fungal genomic DNA and 2 µl of known amounts of the plasmid pAN7-1 were added. The thermal cycler protocol was the following: 95°C for 10 min and 40 cycles with the following steps: 95°C for 30 s; 55°C for 30 s and 72°C for 45 s. Epi-fluorescence microscopic analysis GFP-labelled F. fujikuroi mycelium and spores, grown on PDA plates or inoculated roots, were photographed using an epifluorescence microscope (Axioscope, Zeiss) equipped with a GFP filter and a camera to capture images of GFP fluorescence (excitation at 455 to 490 nm and emission at 515 to 560 nm). Pathogenicity assay To test pathogenicity and virulence of GFP transformants, compared to the wild type isolates, they were grown on PDA or PDB at 23˚C for conidia production. Conidia were harvested and resuspended in water at the concentration of 106 conidia ml-1. The Italian rice cultivar Galileo was used, based on its susceptibility to Bakanae showed in several experiments performed by our group (Valente et al. 2016; Volante et al. 2017). Rice seeds were surface sterilized in 70% ethanol for 1 min, then in 2% sodium hypochlorite for 3 min and subsequently rinsed 3 times in sterile water, finally were allowed to dry at room temperature before sowing. Thirty-two seeds were inoculated with each fungal isolate 7 by pipetting 2 ml of the conidial suspension to each seed. The seeds of control plants (mock) were treated in the same way but inoculating them with sterile deionized H2O. The inoculated seeds were transplanted into 32-cell plastic pots filled with Radicom soil (Vigorplant). A complete randomized block design with three replicates was used. Plants were kept in the greenhouse at 25–28°C under fluorescent lights, with a 12 h photoperiod. After 30 days, seedlings were evaluated for symptoms. Disease severity was evaluated using a scale from 0 to 4 as described by Zainudin et al. (2008) and modified by Valente et al. (2016). The scale includes 5 classes: 0 = no symptoms; 1 = normal growth but leaves beginning to show yellowish–green and/or small necrotic lesions localized at the crown level; 2 = abnormal growth, elongated, thin and yellowish-green leaves, stunted seedlings, necrotic lesions on main root and crown; 3 = abnormal growth, elongated stems, chlorotic, thin and brownish leaves, larger leaf angle, seedlings also shorter or taller than normal, reduced root system with necrotic lesions on secondary roots and on basal stem; 4 = dead plants before or after emergence. One or more of the described symptoms, for each class, could be present on the infected plants. Evaluation of virulence of the isolates was performed as described in Scherm et al. (2019) and infection severity was calculated by the McKinney index (McKinney 1923), here named disease index (DI). Analysis of variance (ANOVA) was performed using COSTAT (version 6.311.; CoHort Software, Monterey, CA, USA) to evaluate the McKinney index data. Data were arcsine-transformed prior to ANOVA analysis. The means were separated using Student–Newman–Keuls multiple-range tests (P < 0.05). Root infection assay and confocal microscopy analysis Two rice varieties, the bakanae disease resistant japonica variety Selenio and the susceptible japonica variety Galileo, were used in this study. Selenio was selected as one of the most resistant rice cultivars from 138 diverse Italian rice accessions screened for evaluation of rice bakanae disease resistance (Volante et al. 2017). Seeds were surface sterilized in 2% NaOCl for 2 min and rinsed in sterile H2O before plating on sterile wet paper for germination. After 5 days at 30°C in the dark, the young emerged roots were inoculated by the Ff 297-GFP strain, by applying 100 µl of a suspension at 106 8 spores ml−1 in the middle of the rootlets, seedlings were allowed continuing the growth at 30°C in the dark until confocal laser scanning microscopic analysis or chitinase expression analysis. After 48 hours after inoculation (hai), 72 hai and 8 days after inoculation (dai), infected rice roots were stained with propidium iodide (0.2 μg ml−1) for 3 min before microscope observation, both unaltered and hand-sectioned roots were analysed. Images of GFP-labelled F. fujikuroi strain in host roots were captured using a confocal laser scanning microscope FV1000 Olympus (Tokyo, Japan) equipped with inverted microscope IX 81. Images were acquired in z stack with objective 10x (N.A. 0,40), using 488nm (argon Ion, emission 520nm) for GFP fluorescence, and 543nm (HeNe, emission 570 nm) laser for propidium iodide staining of root bark. Subsequently they were processed using Imaris 6.2.1 software (Bitplane, Switzerland). Expression analysis of rice Chitinase 1 Total RNA was extracted using the RNeasy kit (Qiagen, Germany) from root tissues (0.1 g) at 72 hai with the selected GFP-tagged F. fujikuroi strain Ff 297-GFP. RNA was treated with TURBO DNA- free kit to remove contaminating DNA (Ambion, Foster City, California, United States). The absence of DNA contamination in RNA samples was further assessed by PCR using the rice elongation factor 1-alpha gene (Manosalva et al. 2009). Total RNA was quantified by Nanodrop (Thermo Fisher Scientific, Waltham, Massachusetts, United States). Reverse transcription reaction was performed using the iScript cDNA synthesis kit (Biorad, Hercules, California, United States). cDNA was then used for expression analysis by quantitative PCR (Applied Biosystem StepOnePlus, Foster City, California, United States) using the specific primers CHIT1-FW (TACTCGTGGGGCTACTGCTT) and CHIT1-RV (CGGGCCGTAGTTGTAGTTGT) for the quantification of the Chitinase 1 rice gene. The primers were designed using the Primer3Plus software (Untergasser et al. 2007). The PCR mix was composed of 5 µl of SYBR Green Power Mix (Applied Biosystem), 2 µl of cDNA, 0.15 µl of each primer (10 µM each) and 2.4 µl of nuclease free water. The thermal cycler protocol was the following: 95°C for 10 min, followed by 40 cycles (95°C for 15 s; 60°C for 60 s) and 95°C for 15 s. 9 The rice elongation factor 1-alpha was used as housekeeping gene with primers EF1α1F and EF1α1R (Manosalva et al. 2009), listed in Online Resource 2. The efficiency of the primers was tested with a standard curve built upon five serial dilutions (1:10) in three technical replicates. After calculating the fold change values, significant differential expression was evaluated with the Duncan’s Post Hoc test, using SPSS v.25. RESULTS AND DISCUSSION Development of gfp-expressing Fusarium fujikuroi isolates Four different F. fujikuroi isolates, named Ff 192, Ff 297, Ff 364 and Ff 1550, were transformed with the vector pCAMgfp introduced into the A. tumefaciens AGL-1 strain. Up to now, F. fujikuroi transformation methods have all been based on the use of protoplasts (Watanabe et al. 2007; Kato et al. 2012; Lee et al. 2018; Chen et al. 2020). However, protoplast production is time consuming and, even for the same isolates, strictly dependent on the batch of lysing enzymes used. Since several years, the Agrobacterium tumefaciens-mediated transformation (ATMT) systems successfully overcame the protoplast-based ones in fungi. Moreover, ATMT-based methods facilitate vector DNA integration in a single site of the recipient genome, and are applicable at different developmental stages, such as conidia, mycelium and fruiting bodies, but germinating conidia are preferred in most of cases, if available (Michielse et al. 2005; Lakshman et al. 2012). Colony morphology and growth rate of F. fujikuroi transformed isolates were similar of those of wild-type isolates, suggesting that gfp expression did not affect the pathogen phenotype. Online resource 3 showed an image of the Ff 297-GFP and the parental Ff 297 isolates selected for microscopy analyses, grown on PDA plates. Approximately, 80-85% of the transformants showed strong fluorescent signal, furthermore, strong fluorescence could be visualized in fungal spores and mycelium (Figure 1). The fluorescence of GFP in transformed F. fujikuroi strains remained stable through subsequent cultivation onto PDA medium without antibiotic, indicating the stable integration of the transforming 10 plasmid in the genome. The number of pCAMgfp copies integrated into the transformant genomes varied from 1 to 2 in the different isolates, Ff 297-GFP had only one copy (data not shown). Pathogenicity of GFP transformants All the four gfp-expressing isolates were found to be pathogenic in infection assays of seeds, but they showed different virulence (Table 1). Ff 192 WT and Ff 297 WT were the most virulent (disease index, DI = 68.0 and 60.0, respectively). Considering the corresponding transformants, Ff 192-GFP DI was 28.0, suggesting that, in this isolate, sgfp gene insertion affected fungal virulence, in a direct or indirect way. On the contrary, Ff 297 WT virulence was not, or little, affected in the corresponding GFP-tagged isolate (DI = 60.0 and 47.0, respectively) and no statistical differences were found between the Disease Indices from Ff 297 WT and Ff 297-GFP. Ff 364-GFP and Ff 1550-GFP showed virulence comparable to parental isolates, but the DI values were lower than Ff 297-GFP (Table 1), so that Ff 297-GFP was selected for root infection and subsequent microscopic analyses. In figure 2A the phenotype of Ff 297-GFP and of the parental strain is illustrated. We already mentioned that F. fujikuroi causes different symptoms on rice, as pre-emergence seedling death, elongated and thinner leaves, chlorosis, stunting, crown rot and root rot and even death of seedlings (Ou 1985; Sunani et al. 2019; Piombo et al. 2020). In figure 2A and B more than one of these symptoms are visible in the seedlings inoculated by Ff 297-GFP and Ff 297: the number of plants is lower than in the mock test (T, on the left), indicating a pre-emergence seedling death; many leaves and stems are elongated and thinner than in the control and have a larger leaf angle; some of them show stunted growth. Infection of susceptible and resistant rice cultivars with one gfp-expressing F. fujikuroi strain The two cultivars, Galileo and Selenio, were previously tested for their response to seed inoculation by F. fujikuroi, showing a susceptible and resistant profile, respectively (Matic et al. 2016; Siciliano et al. 2015; Volante et al. 2017). In our experience and as confirmed by Chen et al. (2020), F. fujikuroi 11 was always detected in the basal roots, so we chose this tissue as the target for pathogen infection and investigation of direct interaction between F. fujikuroi and rice. In this study, a virulent F. fujikuroi strain constitutively expressing the gfp reporter gene was obtained via Agrobacterium transformation of conidia, enabling us to study the early stages of F. fujikuroi infection of rice roots in one resistant and one susceptible cultivar. The infection process in root tissues of the rice cultivars Galileo and Selenio was followed by using the GFP-tagged F. fujikuroi strain Ff 297-GFP and visualized by confocal microscopy at 24 hai, 48 hai, 72 hai, 96 hai, 8 dai and 12 dai of roots. In figure 3, the images captured at 48 and 72 hai were showed, which better illustrated the differences between susceptible and resistant cultivars. Hyphae growing longitudinally along the root surface and in the root hairs were primarily observed (48 hai), and penetration into the epidermal root cells was clearly observed at 72 hai. By this time, most epidermal cells were invaded by the fungus in the susceptible cv Galileo. A similar pattern of hyphal colonization was observed in the roots of the resistant cultivar Selenio, although host cell colonization was much lower in Selenio than that on Galileo (Figure 3). Confocal imaging of transverse sections of the roots showed that the fungus penetrated the stele in both varieties and was more abundant in the susceptible variety Galileo than in the resistant one, Selenio. These results are in agreement with the observations of Lee et al. (2018) at the basal stem level of susceptible and resistant rice cultivars: after ten days from seed inoculation by F. fujikuroi the pathogen was present in the stele and was more abundant in the susceptible one. We cannot exclude that this evidence was due to the major amount of fungal biomass in the susceptible cultivar. However, in literature no significant differences of the amount of F. fujikuroi, measured by qPCR, were reported between the roots of the resistant cultivar Selenio and the susceptible cultivar Dorella (Carneiro et al. 2017). Moreover, Chen et al. (2020) and Cheng et al. (2020) observed, in other rice cultivars, that F. fujikuroi biomasses were similar, both in the stem and roots, in resistant and susceptible cultivars. Confocal analysis of root transverse sections showed colonization of the xylem vessels in both genotypes at 72 hai. The fluorescence was restricted to the vessels in Selenio, while in Galileo there was a more generalized labelling around the vessels. In other 12 words, fungal colonization occurred in both the cultivars, though the fungal presence was less abundant in the resistant one (Figure 3). At 8 dai the roots were completely covered by the fungal hyphae and the diffuse fluorescence did not allow any microscopic analysis (data not shown). This confirmed that fluorescence of F. fujikuroi transformants was stable over time even inside the host tissues. In conclusion, F. fujikuroi is able to colonize the root tissues of both varieties, though Selenio proves to be resistant when seeds are inoculated. These results are in agreement with that observed by Carneiro et al. (2017), which detected the pathogen by qRT-PCR analysis on both the roots of susceptible and resistant cultivars after seed artificial inoculation. Other phytopathogenic Fusarium spp., such as a Fusarium oxysporum f.sp. cubense race 4 isolate tagged by GFP, showed the capacity of invading epidermal cells of host roots directly, and spores were produced in the root system. However, in this case, roots of susceptible banana plants were colonized, but not those of the resistant cultivar, probably due to the production of host exudates that inhibited the germination and growth of pathogenic isolate (Li et al. 2011). We observed that the earliest infectious structures were represented by the infection hyphae, as recently reported by Sunani et al. (2019) by scanning electron microscope analysis. The infection hyphae penetrated the epidermal cells of rice roots at 48-72 hai, and at those times the mycelium was found intra and intercellularly and was able to colonize the vascular bundles. Intercellular and intracellular growth in roots has been documented for other phytopathogenic Fusarium spp., including F. culmorum on rye root tissue (Jaroszuk-Ściseł et al. 2008) and F. oxysporum f. sp. radicis lycopersici on tomato (Lagopodi et al. 2002). Until now, several studies on the rice-F. fujikuroi interaction have been carried out on the aerial part of plants and at several weeks after inoculation (Ji et al. 2016; Ji et al. 2019; Matić et al. 2016). Very recently, the process of infection by F. fujikuroi in rice seeds, seedlings and adult plants was documented in the literature (Sunani et al. 2019; Chen et al. 2020). In the present study we evaluated, for the first time, the progress of infection and colonization in rice roots after direct inoculation of roots by a GFP-tagged isolate of F. fujikuroi. Sunani et al. (2019) addressed details of root 13 colonization by this pathogen after seedling inoculation of a susceptible cultivar. By scanning electron microscopy, they observed the intra and intercellular movement of mycelium at 24 hai and the fully colonization of root tissue at 72 hai of rice seeds. These observations were confirmed in our work after root inoculation, by using a GFP isolate of F. fujikuroi combined with confocal microscopical analysis. Chen et al. (2020) examined the infection in susceptible and resistant seedlings and adult plants by using a GFP-tagged isolate for seed inoculation. In agreement with that observed by Chen et al. (2020), we found that in root tissues the F. fujikuroi infection process was similar in susceptible and resistant cultivars. The fungus, in the first steps of infection, grew predominantly in the root-hair region and was associated with large production of hyphae which directly penetrated the epidermis. The vascular bundles were already colonized at 72 hai, while Chen et al. (2020) observed the hyphae penetration of exodermis and the central cylinder after 6-8 dai, this discrepancy could be due to the use of confocal microscopy, in our case, which is a more powerful technique. We did not observe any appressorium-like structure, in agreement with that reported by Chen et al. (2020), though appressoria formation was reported in several pathogenic Fusarium species, and Sunani et al. (2019) observed them on rice roots after 12 h after seed inoculation by F. fujikuroi. We could hypothesize that at 24 hai, the first time point we considered in our work, the appressoria had already penetrated the host tissue and mycelium was developed. Chitinase expression analysis Chitinases are proteins involved in the plant defence against pathogens because of their ability to hydrolyse chitin in the cell wall of fungi (Sharma et al. 2011). Previous studies revealed that rice leaves expressed chitinases at 2-3 weeks after seed inoculation by F. fujikuroi (Matic et al. 2016). Based on these evidences, in preliminary experiments we tested expression in roots of three genes: Chitinase 1, Gibberellin 20 oxidase 1 and Peroxidase P7. After inoculation by the four wild type F. fujikuroi isolates, only Chitinase 1 was differentially expressed in inoculated rice roots, compared to uninoculated ones. For this reason, we selected this gene for evaluating the expression with the Ff 14 297-GFP transformed isolate. Results showed that Chitinase 1 was expressed in roots also after root inoculation by the Ff 297-GFP isolate. We observed that Selenio and Galileo expressed Chitinase1 at similar levels in the non-inoculated roots, but the expression increased greatly upon pathogen challenge in the resistant cultivar Selenio (Figure 4). The results were in agreement with those obtained with the wildtype isolate (data not shown), excluding that transformation was responsible of the increased chitinase expression in inoculated roots. It has been suggested that in filamentous fungi, chitinases may act during hyphal growth (Kumar et al. 2018), therefore, the induction of Chitinase1 in Selenio may be involved in the control of hyphal growth during the infection, and correlates well with the phenotype of resistance observed in this cultivar. Up-regulation of Chitinase 1 during incompatible interaction between rice and M. oryzae has also been reported (Kawahara et al. 2012). We cannot draw any conclusion regarding the susceptible cultivar Galileo because standard deviation (SD) values of the fold change were too high in Galileo inoculated samples. We repeated the assay three times and always observed that, at 72 hai, in Galileo many germinated seeds showed reduced root length compared to the same not inoculated cultivar. In conclusion, the Galileo inoculated sample was not homogeneous, and this could be a possible explanation of high SD when analyzing gene expression. The reduced root length observed was in agreement with that reported by Zainudin et al. (2008) after seed inoculation. They included the “reduction of root system” among the symptoms of class 3 of the disease scale we used for the pathogenicity test of transformed and untransformed F. fujikuroi isolates. We strongly believe that expression of more and different genes deserves to be explored in the root system, in future studies. CONCLUSIONS Roots represent the first specialized tissue emerging from seeds upon germination, so it might represent an easy tool to study the early stages and the mechanisms performed for rice infection by the seedborne fungal pathogen F. fujikuroi. A F. fujikuroi isolate, constitutively expressing the gfp reporter gene, was successfully obtained via Agrobacterium transformation of a virulent strain. The 15 GFP isolate did not show significant reduction in virulence and was used for studying the early stages of F. fujikuroi infection of rice roots in both resistant and susceptible rice cultivars, by confocal microscope analysis. Though F. fujikuroi is mainly a seedborne pathogen, we found several analogies in infection and colonization of roots after direct root inoculation and seed inoculation, recently well documented in literature. In both cases, the infection process is similar in susceptible and resistant cultivars. The stele is colonized by the pathogen at both basal stem and root level and F. fujikuroi is able to spread both in resistant and susceptible rice plants, although a reduction in fungal colonization in the resistant variety has been observed. This suggests that F. fujikuroi is able to survive and grow inside root tissue even when not causing symptoms. Overall, these results suggest that visualization of F. fujikuroi in roots will help in investigating the early stages of the infection process by this fungal pathogen in rice, while representing a useful tool for the screening of rice cultivars for resistance/susceptibility to F. fujikuroi. Further research is in progress to evaluate and compare the behaviour of F. fujikuroi within the seeds of susceptible and resistant varieties after artificially inoculations, by a GFP-tagged isolate, of floret, which represents the main route of entry of this pathogen. Figure captions Fig. 1. Morphological characteristics of transformed isolates of F. fujikuroi. (A) Typical growth of gfp-expressing F. fujikuroi isolates; (B, C, D, E) Fluorescent and transmission micrographs of gfp- expressing F. fujikuroi spores; (F) Confocal image of fluorescent mycelium on PDA plates; (G) Epifluorescent image of mycelium on the surface of a rice seed. Bars: 20 µm (B, C, D, F), 10 µm (E). Fig. 2. (A) Phenotype of the rice susceptible variety Galileo at 30 day with the gfp-expressing F. fujikuroi isolate Ff 297-GFP (in the middle) and wild type Ff 297 (on the right). T, in the left, represents the control mock-inoculated with dH2O. (B) A closer image of symptomatic plants inoculated by Ff 297. Some of the typical symptoms of Bakanae are visible in the inoculated plants: 16 a lower number of plants, indicating pre-emergence seedling death; many leaves and stems elongated and thinner than in the control and with a larger leaf angle; stunted growth. Fig. 3. Rootlets of rice cv. Galileo (susceptible) and Selenio (resistant), inoculated with the gfp- expressing F. fujikuroi isolate Ff 297-GFP. Root surface and transverse sections at the indicated time after inoculation are shown. Bars: 70 µm for transverse sections, 150 µm for the other pictures. Fig. 4. Expression of Chitinase1 gene in the rootlets of resistant (Selenio) and susceptible (Galileo) rice cultivars. Data obtained by reverse transcriptase real time PCR. The error bar is the standard deviation, and the letters indicate groups not considered to be statistically different using the Duncan test. Electronic Supplementary Material Online Resource 1. Fusarium fujikuroi strains used in this study. Online Resource 2. Primers used in this study. Online Resource 3. Phenotype of the GFP-tagged isolate Ff 297-GFP selected for microscopical analyses and the parental strain Ff 297, both grown on PDA plates. Compliance with Ethical Standards: There are no potential conflicts of interest. This research is not involving human participants and/or animals, therefore, there is no informed consent needed. All the authors have been informed and consent to publish this work. REFERENCES Amatulli, M. T., Spadaro, D., Gullino, M. L., & Garibaldi, A. (2010). Molecular identification of Fusarium spp. associated with bakanae disease of rice in Italy and assessment of their pathogenicity. Plant Pathology, 59(5), 839-844. 17 Campos-Soriano, L. & San Segundo, B. (2009). Assessment of blast disease resistance in transgenic PRms rice using a gfp-expressing Magnaporthe oryzae strain. Plant Pathology, 58, 677–689. Campos-Soriano, L., Valè, G., Lupotto, E. & San Segundo, B. (2013). Investigation of rice blast development in susceptible and resistant rice cultivars using a gfp-expressing Magnaporthe oryzae isolate. Plant Pathology 62, 1030–1037. Carneiro, A.G., Matic, S., Ortu, G., Garibaldi, A., Spadaro, D., & Gullino, M.L. (2017) Development and validation of a TaqMan real time PCR assay for the specific detection and quantification of Fusarium fujikuroi in rice plants and seeds. Phytopathology 107, 885-892. Carter, L. L. A., Leslie, J. F., & Webster, R. K. (2008). Population structure of Fusarium fujikuroi from California rice and water grass. Phytopathology, 98(9), 992-998. Cen, Y-K., Lin, J-G., Wang, Y-L., Wang, J-Y., Liu, Z-Q. & Zheng, Y-G. (2020). The gibberellin producer Fusarium fujikuroi: methods and technologies in the current toolkit. Frontiers in Bioengineering and Biotechnology, 8,232. Chen, C-Y., Chen, S-Y., Liu, C-W., Wu, D-H., Kuo, C-C., Lin, C-C., et al. (2020). Invasion and colonization pattern of Fusarium fujikuroi in rice. Phytopathology, 110 (12). https://doi.org/10.1094/PHYTO-03-20-0068-R Chen, S-Y., Lai, M-H., Tung, C-W., Wu, D-H., Chang, F-Y., Lin, T-C. & Chung, C-L. (2019). Genome-wide association mapping of gene loci affecting disease resistance in the rice-Fusarium fujikuroi pathosystem. Rice 12, 85. Chen, Y. C., Lai, M. H., Wu, C. Y., Lin, T. C., Cheng, A. H., Yang, C. C., et al. (2016). The genetic structure, virulence, and fungicide sensitivity of Fusarium fujikuroi in Taiwan. Phytopathology, 106 (6), 624-635. Cheng, A-P., Chen, S-Y., Lai, M-H., Wu, D-H., Lin, S-S, Chen, C-Y. & Chung, C-L. (2020). Transcriptome analysis of early defenses in rice against Fusarium fujikuroi. Rice 13,65. Chiu, W., Niwa, Y., Zeng, W., Hirano, T., Kobayashi, H. & Sheen, J. (1996). Engineered GFP as a vital reporter in plants. Current Biology, 6, 325–330. Desjardins, A. E., Manandhar, H. K., Plattner, R. D., Manandhar, G. G., Poling, S. M., & Maragos, C. M. (2000). Fusarium species from Nepalese rice and production of mycotoxins and gibberellic acid by selected species. Applied and Environmental Microbiology, 66(3), 1020-1025. Frandsen, R. J. N., Schütt, C., Lund, B.W., Staerk, D., Nielsen, J., Olsson, S. & Giese, H. (2011). Two novel classes of enzymes are required for the biosynthesis of aurofusarin in Fusarium graminearum. The Journal of Biological Chemistry, 286 (12), 10419 –10428. Hossain, M.S., Ayub, A.M., Taher Mia, MA., Islam, M.S., Moni, Z.R. (2013). Estimation of yield loss by Fusarium moniliforme caused bakanae disease of rice. Eco-friendly Agriculture Journal, 6, 40–3. Infantino, A., Balmas, V., Scherm, B., Orzali, L., Ireland, K.B., Laurence, M.H., et al. (2017). First report of Fusarium fujikuroi in the Lao PDR. Australasian Plant Disease Notes,12,14. https://doi.org/10.1007/s13314-017- 0238-5 Ito, S., Kimura, J. (1931). Studies on the ‘bakanae’ disease of the rice plant. Report of Hokkaido National Agriculture Experimental Station, 27, 1-95. 18 Jaroszuk-Ściseł, J., Kurek, E., Winiarczyk, K., Baturo, A. & Łukanowski, A. (2008). Colonization of root tissues and protection against Fusarium wilt of rye (Secale cereale) by nonpathogenic rhizosphere strains of Fusarium culmorum. Biological Control 45(3), 297–307. Jeon, Y. A., Yu, S. H., Lee, Y. Y., Park, H. J., Lee, S., Sung, J. S., et al. (2013). Incidence, molecular characteristics and pathogenicity of Gibberella fujikuroi species complex associated with rice seeds from Asian countries. Mycobiology, 41(4), 225-233. Ji, Z., Zeng, Y., Liang, Y., Qian, Q., & Yang, C. (2016). Transcriptomic dissection of the rice– Fusarium fujikuroi interaction by RNA-Seq. Euphytica, 211(1), 123-137. Ji, Z., Zeng, Y., Liang, Y., Qian, Q., & Yang, C. (2019). Proteomic dissection of the rice-Fusarium fujikuroi interaction and the correlation between the proteome and transcriptome under disease stress. BMC Genomics, 20(1), 91. Kanjanasoon, P. (1965). Studies on the bakanae disease of rice in Thailand. Doc Agr Thesis, Tokyo University, Japan. Karov, I., Mitrev, S., Mihajlov, L., Ristova, D., Arsov, E., & Kovacevik, B. (2005). Gibberella fujikuroi (Sawada) Wollenweber, the new parasitical fungus on rice in region of Kocani. Yearbook, Institute of Southern Crops Strumica, 157-162. Karov, I., Mitrev, S., & Arsov, E. (2009). Gibberella fujikuroi (Wollenweber) the new parasitical fungus on rice in the Republic of Macedonia. Proceedings of National Science Matica Srpska Novi Sad, (116), 175-182. Kato, A., Miyake, T., Nishigata, K., Tateishi, H., Teraoka, T. & Arie, T. (2012). Use of fluorescent proteins to visualize interactions between the Bakanae disease pathogen Gibberella fujikuroi and the biocontrol agent Talaromyces sp. KNB-422. Journal of General Plant Pathology 78, 54–61. Kawahara, Y., Oono, Y., Kanamori, H., Matsumoto, T., Itoh, T., & Minami, E. (2012). Simultaneous RNA-seq analysis of a mixed transcriptome of rice and blast fungus interaction. PLoS One, 7:e49423. Khan, J. A., Jamil, F. F., & Gill, M. A. (2000). Screening of rice varieties/lines against bakanae and bacterial leaf blight (BLB). Pakistan Journal of Phytopathology, 12(1), 6-11. Kim, B. R., Han, K. S., Hahm, S. S., Kwon, M. K., & Nam, Y. G. (2015). Occurrence of the rice bakanae disease in Chungnam province. Research in Plant Disease, 21, 154. Kumar, M., Brar, A., Yadav, M., Chawade, A., Vivekanand, V. & Pareek, N. (2018). Chitinases: Potential Candidates for Enhanced Plant Resistance towards Fungal Pathogens. Agriculture, 8, 88. Kumar, P., Sunder, S., & Singh, R. (2015) Survival of Fusarium moniliforme causing foot rot and bakanae disease in different parts of rice grains. Indian Phytopathology, 68, 454–455. Lagopodi, A. L., Ram, A. F. J., Lamers, G. E. M., Punt, P. J., Van den Hondel, C. A. M. J. J., Lugtenberg, B. J. J., & Bloemberg, G.V. (2002). Novel aspects of tomato root colonization and infection by Fusarium oxysporum f. sp. radicis-lycopersici revealed by confocal laser scanning microscopic analysis using the green fluorescent protein as a marker. Molecular Plant Microbe Interaction, 15(2), 172–179. 19 Lakshman, D.K., Pandey, R., Kamo, K., Bauchan, G., & Mitra, A. (2012). Genetic transformation of Fusarium oxysporum f.sp. gladioli with Agrobacterium to study pathogenesis in Gladiolus. European Journal of Plant Pathology, 133, 729-738. Lee, S-B., Hur, Y-J, Cho, J-H., Lee J-H., Kim, T-H., Cho, S-M., et al. (2018). Molecular mapping of qBK1WD, a major QTL for bakanae disease resistance in rice. Rice, https://doi 10.1186/s12284-017- 0197-7 Li, C., Chen, S., Zuo, C. et al. (2011). The use of GFP-transformed isolates to study infection of banana with Fusarium oxysporum f. sp. cubense race 4. European Journal of Plant Pathology 131, 327–340. https://doi.org/10.1007/s10658-011-9811-5 Lorang, J. M., Tuori, R. P., Martinez, J. P., Sawyer, T. L., Redman, R. S., Rollinset, J. A. et al. (2001). Green Fluorescent Protein Is Lighting Up Fungal Biology. Applied and Environmental Microbiology, 67, 1987–1994. Manosalva, P. M., Davidson, R. M., Liu, B., Zhu, X., Hulbert, S. H., Leung, H. & Leach, J. E. (2009). A germin-like protein gene family functions as a complex quantitative trait locus conferring broad- spectrum disease resistance in rice. Plant Physiology, 149, 286-296. Matić, S., Bagnaresi, P., Biselli, C., Carneiro, G. A., Siciliano, I., Valé, G., et al. (2016). Comparative transcriptome profiling of resistant and susceptible rice genotypes in response to the seedborne pathogen Fusarium fujikuroi. BMC Genomics, 17, 608. McKinney, H. H. (1923). Influence of soil temperature and moisture on infection of wheat seedlings by Helminthosporium sativum. Journal of Agricultural Research, 26, 195-217 Michielse, C.B., Hooykaas, P. J. J., van den Hondel, C. A., & Ram, A. F. J. (2005). Agrobacterium- mediated transformation as a tool for functional genomics in fungi. Current Genetics, 48(1),1-17. Oren, L., Ezrati, S., Cohen, D., & Sharon, A. (2003). Early events in the Fusarium verticillioides- maize interaction characterized by using a green fluorescent protein-expressing transgenic isolate Applied and Environmental Microbiology, 69,1695–701. Ou, S.H. (1985). Bakanae disease and foot rot. In: Rice diseases survey. Kew: Commonwealth Mycological Institute, pp 262–272. Piombo, E., Bosio, P., Acquadro, A., Abbruscato, P. & Spadaro, D. (2020). Different Phenotypes, Similar Genomes: Three Newly Sequenced Fusarium fujikuroi Strains Induce Different Symptoms in Rice Depending on Temperature. Phytopathology 110(3):656-665. doi: 10.1094/PHYTO-09-19- 0359-R. Saremi, H., Ammarellou, A., Marefat, A., & Okhovvat, S. M. (2008). Binam a rice cultivar, resistant for root rot disease on rice caused by Fusarium moniliforme in Northwest, Iran. International Journal of Botany,4, 383-389. Scherm, B., Balmas, V., Infantino, A., Aragona, M., Valente, M.T., Desiderio, F., et al. (2019) Clonality, spatial structure, and pathogenic variation in Fusarium fujikuroi from rain-fed rice in southern Laos. PLoS ONE, 14(12): e0226556. https://doi.org/10.1371/journal.pone.0226556 20 Sesma, A. & Osbourn, A.E. (2004). The rice leaf blast pathogen undergoes developmental processes typical of root-infecting fungi. Nature 431, 582–586. Sharma, N., Sharma, K. P., Gaur, R. K., & Gupta, V. K. (2011). Role of chitinase in plant defense. Asian Journal of Biochemistry,6(1), 29-37. Siciliano, I., Amaral Carneiro, A., Spadaro, D., Garibaldi, A., & Gullino, M.L. (2015). Jasmonic acid, abscisic acid and salicylic acid are involved in the phytoalexin responses of rice to Fusarium fujikuroi, a high gibberellin producer pathogen. Journal of Agricultural and Food Chemistry 63, 8134-8142. Sunani, S. K., Bashyal, B. M., Kharayat, B. S., Prakash, G., Krishnan, S. G., & Aggarwal, R. (2019). Identification of rice seed infection routes of Fusarium fujikuroi inciting bakanae disease of rice. Journal of Plant Pathology, https://doi.org/10.1007/s42161-019-00390-8. Untergasser, A., Nijveen, H., Rao, X., Bisseling, T., Geurts, R., & Leunissen, J. A. (2007). Primer3Plus, an enhanced web interface to Primer3. Nucleic acids research, 35(suppl_2), W71-W74. Valente, M. T., Desiderio, F., Infantino, A., Valè, G., Abbruscato, P. & Aragona, M. (2016). Genetic variability of Fusarium fujikuroi populations associated with Bakanae of rice in Italy. Plant Pathology, https://doi: 10.1111/ppa.12575. Volante, A., Tondelli, A., Aragona, M., Valente, M.T., Biselli, C., Desiderio, F., et al. (2017). Identification of bakanae disease resistance loci in japonica rice through genome wide association study. Rice, 10: 29. Watanabe, S., Kumakura, K., Izawa, N., Nagayama, K., Mitachi, T., Kanamori, M., et al. (2007). Mode of action of Trichoderma asperellum SKT-1, a biocontrol agent against Gibberella fujikuroi. Journal of Pesticide Science, 32(3), 222–228. Wiemann, P., Sieber, C.M.K., von Bargen, K.W., Studt, L., Niehaus, E-M., Espino, J.J., et al. (2013). Deciphering the Cryptic Genome: Genome-wide Analyses of the Rice Pathogen Fusarium fujikuroi Reveal Complex Regulation of Secondary Metabolism and Novel Metabolites. PLoS Pathogens 9(6): e1003475. doi:10.1371/ journal.ppat.1003475 Zainudin, N. I. M., Razak, A., & Salleh, B. (2008). Bakanae disease of rice in Malaysia and Indonesia: etiology of the causal agent based on morphological, physiological and pathogenicity characteristics. Journal of Plant Protection Research, 48(4), 475-485. isolate McKinney index Ff 192 WT 68.0 ± 9.75 a Ff 192 GFP 28.0 ± 9.75 b Ff 297WT 60.0 ± 15.0 a Ff 297GFP 47.0 ± 6.71 ab Ff 364 WT 39.0 ± 17.82 b Ff 364 GFP 24.0 ± 15.57 b Ff 1550 WT 27.0 ± 13.51 b Ff 1550 GFP 31.0 ± 5.48 b Table 1. Comparison of virulence between the wild type and GFP transformed F. fujikuroi isolates. Means ± Standard deviations followed by different letters are significantly different (P≤ 0.05), according to Student-Newman-Keuls test. 21 Figure 1 F G Figure 2 A B Figure 3 Galileo: 48 hpi 72 hpi 72 hpi transverse section Selenio: 48 hpi 72 hpi 72 hpi transverse section Figure 4