Fungal sclerotia (ca. 2 mm diameter, see Figure 1B) were collected from a symptomatic colony of R. japonicum and sterilized with sodium hypochlorite solution (8.5 to 13.5% active chlorine; Nakalai Tesque, Kyoto, Japan) diluted 10-fold [0.05% (w/w) Tween 20] for 2 min, followed by five rinses with sterilized water. Sclerotia were cultured on potato dextrose agar (PDA) in Petri dishes under continuous light-emitting diode (LED) illumination (mixture of white and red) of ca. 80 μmol m^-2 s^-1 at 23°C. The hyphae that developed from a sclerotium were selected and subcultured to establish a single fungal isolate, SR1, used for further study. The fungus was maintained on PDA medium (Nissui Pharmaceutical, Tokyo, Japan) or potato sucrose agar (PSA) containing 3.9 g L^-1 potato extract (Formedium, Hunstanton, United Kingdom), 21 g L^-1 sucrose and 14.1 g L^-1 agar under continuous light and the conditions described above.

In our previous study, the peroxidase-encoding gene Prx34 was knocked out in P. patens (Lehtonen et al., 2009). The wild type and a Prx34-deficient strain of P. patens (Prx34-KO) were aseptically used. An aseptic culture of M. polymorpha was kindly provided by Dr. K. T. Yamato, Kindai University, Wakayama, Japan. Gametophytes of the mosses R. japonicum and Hypnum plumaeforme Wilson were collected from the campus of the Faculty of Biology-Oriented Science and Technology, Kindai University, Wakayama, Japan. The collected gametophytes were washed in tap water for 60 min prior to use. Seeds of cultivated dicotyledonous plants, comprising Hordeum vulgare L., Brassica rapa L. var. pekinensis (Lour.) Kitam., L. sativa L., and S. oleracea L., were purchased from a local market and surface-sterilized with sodium hypochlorite for 10 min, followed by three rinses with sterilized water. The seeds were germinated in autoclaved Plant Boxes (60 mm × 60 mm × 100 mm; AGC Techno Glass, Shizuoka, Japan) filled with vermiculite (depth 3 cm) (size 3–6 mm, Komeri, Niigata) and 55 mL of commercial liquid fertilizer (Hyponex 6-10-5, Hyponex Japan, Osaka, Japan) diluted 1000-fold. Plants were cultured under continuous LED illumination (mixture of white and red) of ca. 80 μmol m^-2 s^-1 at 23°C.

Protonemata of P. patens were cultured on BCD plate medium (1 mM MgSO₄, 1.85 mM KH₂PO₄, 10 mM KNO₃, 45 mM FeSO₄, 1 mM CaCl₂, 0.22 mM CuSO₄, 0.19 mM ZnSO₄, 10 mM H₃BO₄, 0.10 mM Na₂MoO₄, 2 mM MnCl₂, 0.23 mM CoCl₂, and 0.17 mM KI, solidified with 8 g L^-1 agar) supplemented with 5 mM ammonium tartrate¹. Protonemata (ca. 1 g) were homogenized 30 s in 15 mL of sterilized soft agar (1.5 g L^-1 in water) using a handy homogenizer (Polytron PT1200E, Kinematica, Luzern) for subculturing. A cellophane sheet (No. 300, Rengo, Osaka) was put on the medium and 3 mL of the homogenate was poured on the plate and subcultured every 2 weeks. In vitro gametophytes of P. patens were developed on BCD plate medium by culture for 3 weeks under continuous illumination from a cold cathode fluorescent lamp with ca. 80 μmol m^-2 s^-1 at 23°C.

Genomic DNA was extracted from fungal hyphae grown in aseptic culture. Hyphae were frozen in liquid nitrogen, ground in a mortar, and the powder was suspended in 300 μL lysis buffer (100 mM Tris–Hc1, pH 7.5, SDS 0.5% w/v, 30 mM EDTA) and boiled for 7 min. Subsequently, 150 μL of 3 M sodium acetate was added, followed by incubation at -30°C in a freezer for 7 min. After centrifugation at 13,000 ×g for 5 min, the supernatant was extracted with an equal volume of phenol:chloroform:isoamyl alcohol (1:1:1, v/v/v) and washed with chloroform, 2-propanol, 70% (v/v) ethanol and then desiccated. Finally, the purified DNA was suspended in TE (10:1) buffer. The rDNA internal transcribed spacer (ITS) region, including ITS1 and ITS2, was amplified by PCR for sequencing using KOD FX Neo polymerase (TOYOBO, Osaka, Japan) in accordance with the manufacturer’s protocol. The primers NSI1 (5^′-GATTGAATGGCTTAGTGAG-3^′) and NLB4 (5^′-GGATTCTCACCCTCTATGA-3^′) (Martin and Rygiewicz, 2005) were used. PCR products were analyzed by agarose gel electrophoresis, extracted from the gel using the QIAEX^® II Gel Extraction Kit (QIAGEN, Velo, Netherlands) and sequenced by SigmaGenosys (Sigma-Aldrich Japan, Tokyo, Japan).

Pathogenicity of the isolated fungus on R. japonicum was tested in Petri dishes (diameter 8.5 cm, height 4 cm) containing 15 g vermiculite and 40 mL of 1000-fold diluted commercial liquid fertilizer (Hyponex 6-10-5). The Petri dishes were autoclaved, the gametophytes were transferred onto the vermiculite, then the Petri dishes were packed into a transparent plastic box (10 cm × 20 cm × 6 cm). To test the pathogenicity on non-vascular plants, gametophytes of H. plumaeforme and P. patens, and thalli of M. polymorpha were transferred to autoclaved Plant Boxes (60 mm × 60 mm × 100 mm) containing vermiculite (depth 3 cm) and 55 mL BCD medium. Fungal hyphae cultured for 2 weeks on PDA plates were cut into ca. 0.5 cm² pieces and placed on the vermiculite at a minimum distance of 1.5 cm from the plants. All Petri dishes and Plant Boxes were incubated under continuous LED illumination of ca. 80 μmol m^-2 s^-1 at 23°C.

To test the pathogenicity of the fungal isolate against the protonema of P. patens, a sclerotium was placed directly on protonemata that had been cultured on BCD medium in a 9 cm Petri dish for about 1 week. To test pathogenicity against the gametophyte of the WT and Prx-KO mutant of P. patens, explants were collected from colonies cultured on BCD medium for 3 weeks. Eight gametophytes per plate were concentrically inoculated, each 1.5 cm from the center of the plate, on BCD medium in a 9 cm Petri dish and the sclerotium was placed at the center of the dish. For the pathogenicity test using the gametophyte colonies, six gametophytes were concentrically placed with each 2.0 cm apart from the center on BCD medium in a 9 cm Petri dish and cultured for 3 weeks. A sclerotium was then placed at the center of the dish. All experiments were carried out under continuous light conditions using a cold cathode fluorescent lamp with light intensity of about 80 μmol m^-2 s^-1 at 23°C.

A cuvette (12 i.d. × 55 mm) was filled with 895 μL of 50 mM sodium acetate buffer (pH 5.0) and a sheet of 1 cm² of the protonemata directly cut from the 2 weeks old culture plate. After incubation for 30 min in the dark at room temperature, 5 μL of 1 mM 2-methyl-6-(4-methoxyphenyl)-3,7–dihydroimidazo(1,2-a)pyrazin-3-one hydrochloride (MCLA; Sigma-Aldrich, St. Louis, United States) and secreted material (100 μL) directly collected by careful pipetting from the fungus was added. Measurement with a luminescence meter (Gene Light GL-200, Microtec, Chiba, Japan) was started immediately. The measurement time was set to 5 s.

The antagonistic bacterial strain R-1 (Supplementary Figure S2) was found by chance in a study for isolating fungi coexisting with moss colonies. The 16S rDNA sequence of R-1 (NCBI accession no. AB847954) showed extremely high similarity to that of Bacillus amyloliquefaciens (99% identity). The bacterium was cultured in LB medium for 19 h at 28°C, collected by centrifugation at 13,000 ×g for 10 min, and resuspended in 10-fold diluted BCD medium at OD₆₀₀ = 0.1. An aliquot (5 μL) of the suspension was inoculated onto gametophytes of P. patens cultured on BCD plates, which were cultured under continuous light (ca. 80 μmol m^-2 s^-1) at 23°C for 1 week. Pure-cultured sclerotia (2 mm in diameter) were set 1.5 cm from single gametophytes or 2 cm from the colony of gametophytes as described in the preceding sections.

Hyphae of the fungal isolate SR1 grew well on PDA and PSA plates, and white or slightly brown sclerotia formed 3 weeks after inoculation (Figure 2A). Microscopic observations showed that the hyphae had clamp connections (Figure 2B). After inoculation of R. japonicum with the pure-cultured hyphae, the gametophytes were severely damaged and sclerotia developed (Figure 3). This result indicated that the fungal isolate, SR1, is a destructive pathogen of R. japonicum. The ITS sequence (LC374375) of SR1 was 100% identical with that of S. delphinii strain CBS672.71 (JN241578.1), which suggested that SR1 is a strain of S. delphinii (syn. S. rolfsii var. delphinii).

Colonization of mosses, liverwort and seed plant species by SR1 was observed in plants grown on vermiculite in Petri dishes or Plant Boxes. The hyphae of pure cultures of SR1 placed on vermiculite without plants grew relatively slowly, i.e., 2–3 mm during the 7-day culture period. In contrast, SR1 hyphal growth was visible at 2 days post-inoculation (dpi) when incubated with non-vascular plants, and plants that the hyphae contacted had withered by 4–6 dpi (Figure 4). All mosses and liverworts inoculated, namely R. japonicum, P. patens, H. plumaeforme, and M. polymorpha, were killed by SR1.

When incubated with vascular plant seedlings, the SR1 hyphae grew and reached the plants at 8–11 dpi (Figure 5), except for cultures with plants of S. oleracea in which the hyphae grew more slowly. The plant tissues colonized by the hyphae turned brown or white and withering was observed at 11–18 dpi (Figure 5A–D). Unlike the other species, the majority of the hyphae (over 90%) formed sclerotia in the boxes containing S. oleracea, although hyphal growth was minimal (Figure 5E). However, hyphae grew from several pieces of inoculants and, once the hyphae reached S. oleracea plants (less than 10% of inoculants), the plant was covered by the hyphae, became discolored, and wilted within 14 dpi (Figure 5D). The hyphae of SR1 grew in the plant tissues (Supplementary Figure S2) and developed sclerotia. These results indicated that SR1 is a multihost colonizer of bryophytes, liverworts, and seed plants. SR1 also quickly grew on protonemata of P. patens (Figure 6A,B). The protonema cells colonized by hyphae were swollen, and droplets of fluid developed on the hyphae (Figure 6C–F). The droplets of secretory fluid were also observed on the pure-cultured hyphae on the agar plate, especially during development of sclerotia, 1 week after inoculation. Subsequently, the droplets gradually disappeared, except those around the sclerotia (see also Figure 2A).

It was difficult to maintain R. japonicum gametophytes under the aseptic culture condition, therefore further analysis was performed using P. patens as a model species. We collected the secretion fluid from SR1 hyphae and examined its effect on ROS generation by P. patens protonemata. The fungal secretion rapidly induced ROS production (Figure 7). Production of ROS can be induced also by chitin or chitosan, which involves the peroxidase Prx34 in P. patens (Lehtonen et al., 2009), and the chitosan-treated Prx34-KO mutant generates only weak ROS signals, as detected by MCLA chemiluminescence (Lehtonen et al., 2012a). In response to application of the SR1 secretory fluid, similar amounts of ROS were detected in wild-type (WT) P. patens and the Prx34-KO mutant (Figure 7).

Physcomitrella patens was inoculated by placement of pure-cultured sclerotia of SR1 at a distance of 1.5 cm from gametophytes of the Prx34-KO mutant and the WT. Hyphal growth was quickly initiated, with hyphae reaching the gametophytes at 3 dpi and quickly covering the plants (Figure 8). The WT plants were totally covered by hyphae by 4–5 dpi. In contrast, the hyphae showed more rapid growth on the Prx34-KO mutant and the gametophytes were totally covered by hyphae by 3–4 dpi.

We previously isolated a bacterium R-1, possibly B. amyloliquefaciens, that shows antagonistic activity against fungi (Supplementary Figure S3). The bacterium R-1 showed antagonistic activity against the fungal isolate SR1 (Figure 9), but inoculation with R-1 showed no influence on P. patens gametophytes (see also Figure 10). When sclerotia were inoculated at 2 cm distance from the colony of the gametophytes, hyphae quickly reached the gametophytes and caused browning on WT plants at 4 dpi (Figure 10), whereas for gametophytes pre-inoculated with R-1, growth of hyphae was distinctly repressed and new sclerotia quickly developed on the inoculated sclerotia. Gametophytes of the Prx34-KO mutant were more quickly damaged by SR1 compared with WT gametophytes, as described above, whereas Prx34-KO gametophytes pre-inoculated with R-1 showed no symptoms and the fungus quickly developed new sclerotia (Figure 10).