Dimorphism in Neopseudocercosporella capsellae, an Emerging Pathogen Causing White Leaf Spot Disease of Brassicas

White leaf spot pathogen: Neopseudocercosporella capsellae causes significant damage to many economically important Brassicaceae crops, including oilseed rape through foliar, stem, and pod lesions under cool and wet conditions. A lack of information on critical aspects of the pathogen’s life cycle limits the development of effective control measures. The presence of single-celled spores along with multi-celled conidia on cotyledons inoculated with multi-celled conidia suggested that the multi-celled conidia were able to form single-celled spores on the host surface. This study was designed to demonstrate N. capsellae morphological plasticity, which allows the shift between a yeast-like single-celled phase and the multi-celled hyphal phase. Separate experiments were designed to illustrate the pathogen’s morphological transformation to single-celled yeast phase from multi-celled hyphae or multi-celled macroconidia in-vitro and in-planta. Results confirmed the ability of N. capsellae to switch between two morphologies (septate hyphae and single-celled yeast phase) on a range of artificial culture media (in-vitro) or in-planta on the host surface before infection occurs. The hyphae-to-yeast transformation occurred through the production of two morphologically distinguishable blastospore (blastoconidia) types (meso-blastospores and micro-blastospores), and arthrospores (arthroconidia).


INTRODUCTION
Dimorphic fungi are capable of assuming two distinct morphologies during their life cycle, switching between a single-celled yeast form and a multi-celled hyphal form (Gauthier, 2015). Dimorphism is a highly synchronized, reversible response to an external stimulus posed by the environment (Lin et al., 2015). The majority of dimorphic fungi belong to the class Ascomycete or their 'Imperfect' relatives (Gauthier, 2015), and morphogenesis is widespread among pathogenic fungi of either mammals, plants, or insects (Gauthier, 2015). Well-known dimorphic phytopathogenic fungi such as Taphrina deformans, the cause of peach leaf curl disease, and Ophiostoma novo-ulmi, the cause of Dutch elm disease, have had major adverse impacts on agriculture or the urban landscape, respectively (Gauthier, 2015).
In dimorphic fungi, yeast phase development occurs usually in response to a change in growth condition and/or nutritional status of the growth medium (Odds and Kerridge, 1985;Lin et al., 2015). Although diverse stimuli, including pheromones, plant lipids, plant hydrophobicity, pH, nitrogen, and quorum sensing can induce the transition of yeast-to-mycelium or vice versa in fungal morphogenesis in plant pathogenic fungi (Orlowski, 1994), several human fungal pathogens are solely thermally dimorphic, where temperature alone induces the morphological transformation (Orlowski, 1994;Gauthier, 2015). The search for internal factors that regulate morphogenesis or external factors that stimulate this dimorphic switch is a rapidly growing area of research, due in part to the availability of novel microscopy and molecular tools in the past decade (Gauthier, 2015). Such studies have revealed that pathogenicity, virulence, and lifecycle of dimorphic fungal pathogens are closely interrelated, and that pathogenicity may be limited to one of the alternative morphologies (Gauthier, 2015;Lin et al., 2015).
Despite being a common Brassica pathogen, important knowledge gaps exist in the N. capsellae-Brassica pathosystem (Gunasinghe et al., 2020), particularly on critical aspects of its epidemiology. This is in part from its very slow growth rate on artificial media (Crossan, 1954) and its secretion of cercosporin, a non-host specific, photo-activated toxin into both artificial media and during the early infection process in host tissues (Gunasinghe et al., 2016b). Air-dispersed ascospores produced by the teleomorph are responsible for the disease initiation and long-distance pathogen dispersal in the United Kingdom . However, elsewhere, most N. capsellae populations are anamorphic and thought to be clonal, where aspects of N. capsellae pathogenicity, disease initiation, spread, and epidemic development are not fully described and/or understood (Gunasinghe et al., 2020). For example, on rapeseed cotyledons, Australian isolates of N. capsellae can be observed as hyphae, as large multi-celled conidia, and what appeared to be singlecelled spores that are able to reproduce by budding (Gunasinghe unpub). This suggests that a morphologic shift between the yeast phase and the hyphal form likely occurs for N. capsellae. Its relevance is that morphologic shift between the yeast phase and the hyphal form in other pathogenic fungi is essential for maintaining pathogenicity (Orlowski, 1994). Further, Lin et al. (2015) noted, a thorough understanding of the life cycle of dimorphic pathogens is required if they are to be effectively controlled. Hence, studies were undertaken to determine if N. capsellae is dimorphic and able to switch between a singlecelled yeast form and a multi-celled hyphal form in-vitro or inplanta on host leaves.

Pathogen Isolates
Three single-spored isolates of N. capsellae (UWA Wln-10, UWA Wlj-3, and UWA Wlra-7), stored as lyophilized ampoules containing the living pathogen preserved on vacuum dried agar cultures (Bazzigher and Kanzler, 1985) at room temperature, were obtained from the University of Western Australia (UWA) culture collection. All three isolates were collected within Western Australia. UWA Wln-10 was from infected leaves of B. napus at Bindoon North, UWA Wlj-3 was from diseased B. juncea leaves at Shenton Park, and UWA Wlra-7 was isolated from infected Raphanus raphanistrum leaves at West Calingiri (Gunasinghe et al., 2016d). Pathogen isolates were revived on Potato Dextrose Agar (PDA -Difco) or Malt Yeast Extract (MEA) plates and then working cultures were maintained as MEA media slants at 4°C. For in-vitro studies, two isolates of genetically different groups (UWA Wlj-3, and UWA Wlra-7) (Gunasinghe et al., 2016a) were used. When multi-celled conidia were needed, susceptible host plants grown in the field were spray inoculated with a mixture of hyphal fragments of all three isolates, and conidia produced on host plants lesions were collected (Gunasinghe et al., 2014).

Preparation of Mycelial Suspension and Multi-Celled Conidial Suspension for In-Vitro and In-Planta Studies
To prepare a mycelial suspension, each of the three N. capsellae isolates (UWA Wln-10, UWA Wlj-3, and UWA Wlra-7) was sub-cultured onto MEA in Petri Plates and grown for three weeks at 20°C. Erlenmeyer flasks (250 mL) containing 150 mL of Malt Extract Broth (MEB) were inoculated with agar plugs from the growing edges of each N. capsellae culture separately and incubated on a rotary platform shaker (Innova ™ 2100, New Brunswick Scientific) at 150 rpm and 20°C. After 3-4 weeks, cultures of all three isolates with abundant mycelial growth were mixed in equal volumes and blended for 5 min (Kambrook Mega Blender, Breville Group Ltd). Mycelial mats in the mixture after blending were removed by filtering through sterile gauze. Concentration of the filtered suspension of mycelial fragments containing 90 µm (± 10 µm) length was adjusted to 4 x 10 6 mL -1 using a hemocytometer and sterile distilled water (Gunasinghe et al., 2014).
To produce multi-celled conidia, field-grown plants were inoculated with mycelial inoculum. Seeds of highly susceptible, B. juncea genotypes Rohini, Vardan, or Prakash (Gunasinghe et al., 2014) were sown (6 seeds per pot) in sequential batches of 9 pots each, seedlings were grown in a controlled environment room (15°C, 12h photoperiod and Photosynthetic Photon Flux Density of 580mmolm -2 s -1 ) for 15 to 20 days before transferring into an experimental field plot (1 x 0.5m) at the University of Western Australia, Crawley, Western Australia, where they were fertilized weekly with Thrive ® (DuluxGroup (Australia) Pty Ltd), a complete nutrient solution dissolved in water. Beginning at approximately four weeks of age, field plants were sprayinoculated weekly with a mixture of mycelial fragments (4 x 10 6 fragments mL -1 ) prepared as described above for three consecutive weeks using a handheld and operated aerosol sprayer in the late afternoon to maximize the period of high humidity to favour infections.
After 20 to 25 days post-inoculation, when disease symptoms became apparent, leaves with typical white leaf spot symptoms were collected. Leaves with lesions were gently washed with sterile deionized water (DI) before, lesions were excised and placed into glass vials containing 20-30mL of sterile DI. Multi-celled conidia were released by vigorous handshaking for four min. The pieces of the leaf lesions were removed from the glass vials with a sterile pair of forceps and the mixture was centrifuged to pellet the conidia. The supernatant was discarded and the conidial pellet was washed twice in sterile DI water and re-suspended in 2 to 3mL of sterile DI water. Multi-celled conidial concentration (4 x 10 6 mL -1 ) was achieved by making adjustments based on hemocytometer counts (Gunasinghe et al., 2016c).

Dimorphism In-Vitro
In-Vitro Blastospore Formation From N. capsellae
For the agar plug method, a mycelial plug (5mm diameter) from the growing edge of the mycelial colony was placed in the middle of a Petri Plate containing PDA, MEA, V-8 juice agar, NDYA, YEDPA, or WA. For the streak plate method, a loop full of mycelial fragments was taken from the growing edge of the colony and dragged over the surface of the plate in a 'zig-zag' formation. There were six single plate replicates for each medium for each inoculation method. Three plates from each media inoculated with each individual inoculation method were incubated at 23°C, and the other three at 37°C, in the dark for a maximum of 4 weeks.
Stereo microscopy, light microscopy (LM), and scanning electron microscopy (SEM) techniques were used to document colony and cell morphologies. Colony morphologies of N. capsellae on different media incubated at the two temperatures were recorded every other day by examining individual plates. Colonies on plates were further prepared to examine with a Nikon-SMZ1500 stereomicroscope, a Nikon Eclipse Ni-U light microscope, or SEM where necessary.
LM: Cells/hyphae collected by scraping the surface of colonies were mounted on microscope slides and stained with 1% cotton blue in lactophenol. Slides were studied and photographed (1000x magnification) with a Nikon Eclipse Ni-U light microscope. SEM: Selected colonies were prepared for SEM for further analysis. Thin agar blocks (2-3 cm) with colonies were cut carefully from the culture plate with a sterile scalpel and placed on a pin mount with a carbon tab (ProSci Tech) on it or in cryovials (2mL). The samples on pin mounts were air-dried in a laminar flow cabinet for five days. Cryotubes with samples were frozen at -80°C (Revco ULT390-5-D, Thermo Scientific) for 24h and then freeze-dried (FD-8 Witeg Germany) for 24h. Dried samples were carefully placed on standard pin mounts with a carbon tab on top. Finally, all samples were sputter-coated with Mini Plasma Sputter Coater (gold: 0.1mm) and visualized and imaged using a Jeol JCM-6000 Versatile Benchtop SEM.
In-Vitro Blastospore Formation From N. capsellae

Multi-Celled Conidia
Two drops (20µl) from multi-celled conidial suspension prepared as described before in the previous section on preparation of multi-celled conidial suspension, were placed on 10 standard glass microscope slides and incubated in a moist chamber. After a 24h incubation period, 1% cotton blue in lactophenol was added to each drop, before examining under an Olympus BX5 microscope. From 5 random slides, a total of 50 fields (10 from each) were imaged with an Olympus DP71 digital photographic system.

Morphological Analysis of In-Vitro Formed Blastospores (Single Spore/Cell) Types
Two measurements (length and width) were taken from more than 400 spores using the software ImageJ 1.53a (Schneider et al., 2012) from digital images taken of two different single-celled blastospore types (micro-blastospores and meso-blastospores) produced from N. capsellae hyphae in pure cultures or multicelled macroconidia suspended in sterile distilled water. All the measurements for smaller bacteria-like spores (microblastospores) were taken from SEM images, and the larger yeast-like spore (meso-blastospore) measurements were taken from a combination of both SEM and LM images.

Host Genotypes
Two genotypes from each of three major oilseed species were selected for in-planta studies. The six susceptible or resistant genotypes used were B. carinata (ATC94129P, highly resistant and UWA#012, very susceptible), B. napus (Hyola 42, resistant and Trilogy, susceptible), and B. juncea (Dune, moderately resistant and Vardan, susceptible) (Gunasinghe et al., 2014).
In-Planta Blastospore Formation From N. capsellae Hyphae Seeds of each genotype were sown in 48 cell seedling trays (30 x 34 cm), four seeds per cell and four replicate cells per genotype. Ten days after sowing, plants were thinned to two per cell. Seedlings were grown in a controlled environment room (15°C, 12h photoperiod, and Photosynthetic Photon Flux Density of 580 mmol m -2 s -1 ). Tenday-old cotyledons of each seedling were inoculated by depositing a single drop (10µl) of the mycelial suspension (4 x 10 6 mL -1 ), prepared as described earlier in the previous section on preparing of mycelial suspension, on each cotyledon lobe. The control was an identical tray of seedlings under the same conditions but inoculated with sterile distilled water. All plants were placed in clear polyethylene bags for 48h to maintain high humidity to promote infection (Brun and Tribodet, 1991). Five cotyledons for each genotype, (4 inoculated and 1 control from each of the 5 cells) were randomly sampled for LM or SEM 48h post-inoculation.

Multi-Celled Conidia
Collection of multi-celled macroconidia from diseased leaves and preparation of macroconidial suspension in sterile distilled water (4 x 10 6 mL -1 ) was done as described in the previous section on preparation of muiti-celled conidia. An experiment identical to mycelial inoculation studies was conducted, but by inoculating cotyledons with a 10µm multi-celled conidial suspension rather than mycelial suspension. After 48h post-incubation, two sets of five cotyledons (four inoculated and one control) were collected separately for LM and SEM.

Sample Preparation for Microscopy and Morphological Analysis of Blastospores (Single Spore/Cell Types) Produced In-Planta
For LM studies, collected cotyledons were decolourized with acetic acid: ethanol: water (2:2:1) solution and stained with 1% cotton blue in lactophenol for 40 sec to prepare whole wet mounts (Gunasinghe et al., 2016c). Four inoculated and one uninoculated cotyledons were examined using an Olympus (BX51) microscope, where 25 random images from each inoculated site (50 from a cotyledon and a total of 50 x 4 = 200 for each genotype) were captured using an Olympus DP71 digital photographic system.
For SEM studies, cotyledon samples were fixed in 2.5% glutaraldehyde and cut into 4 x 4 mm pieces and transferred into glass vials with fresh 2.5% glutaraldehyde. Samples were then prepared for SEM using a BioWave microwave processor fitted with a PELCO coldspot (Gunasinghe et al., 2016c). Prepared samples were mounted on standard aluminium pin mounts with carbon tabs (ProSciTech), sputter-coated with 5 nm carbon, 3 nm platinum, and surface imaged with a field emission SEM (Zeiss Supra 55 VP). Twenty-five random images were taken from each of the inoculated areas (50 from a cotyledon and a total of 50x4 = 200 for each genotype).
Different blastospore types formed from N. capsellae multicelled macroconidia or hyphal fragments were measured using the software ImageJ 1.53a. Two measurements from more than 400 micro-blastospores (bacteria-like single-celled spores) from SEM images and meso-blastospores (Yeast-like single-celled spores) from a combination of both SEM and LM images were taken.

Molecular Identification
Confi rmation of the identity of the two observed morphologically different phases (Hyphae and single-celled blastospores) was undertaken by sequencing the internal transcribed spacer region (ITS1, 5.8S, and ITS4) of the rDNA operon, and comparing sequence data with available sequence information data for N. capsellae isolates in GenBank. Areal hyphae obtained from a pure mycelial colony and single-celled blastospores collected from a young bacterial-like colony consisted only of blastospores have been used to extract DNA. To obtain the hyphal phase, five Petri Plates containing 15mL NDY were inoculated with a mycelial plug taken from a threeweek-old culture of the isolate UWA Wlj-3 on MEA plates at the mycelial phase. After 20 to 30 days of incubation at 23°C, plates were examined to select a plate with mycelial growth and no blastospores. Aerial mycelium (hyphae) from selected NDY cultures was harvested and placed in a sterile Eppendorf tube. Streak plates of YEPD (in duplicate) from the same isolate: UWA Wlj-3 were prepared from the same mycelial colony on MEA to obtain yeast phase (single cells). YEDP was used to induce the unicellular phase of the pathogen (Zhu et al., 2017). After two days of incubation at 23°C in dark, cells from the growing yeastlike colonies were aseptically transferred into a sterile Eppendorf tube.

Extraction of Genomic DNA, PCR, and Sequencing
Genomic DNA was extracted using the ISOLATE II Genomic DNA Kit (Bioline (Aust) Pty Ltd) with some modifications to the standard protocol. All samples with mycelia or single cells (yeast phase) were washed with 400µl of TE buffer in an Eppendorf tube, pelleted by centrifugation (13,000rpm, 5min) in a microfuge ® 18 (Beckman Coulter ™ ) and the supernatant was discarded. Lysis buffer (200µl) and two metal beads were added to each of the sample and homogenized with the tissue homogenizer (FastPrep-24TM 5G Homogenizer -MP Biomedicals) for 6 seconds. Genomic DNA extraction in the homogenized samples was carried out using the ISOLATE II Genomic DNA Kit according to manufacturer instructions. All DNA samples were stored at 4°C after evaluating the quality and concentration of DNA extracted using a DeNovix: DS-11 FX Spectrophotometer/Fluorometer. Part of the nuclear rRNA operon was amplified with primers ITS1 (5'TTT CCG TAG GTG AAC CTG C3') and ITS4 (5'TCC TCC GCT TAT TGA TAT GC3') using the polymerase chain reaction (PCR) (Gunasinghe et al., 2016a). Spacer I, between the 18S and 5.8S rRNA gene, was amplified using the 5' primer ITS1, and spacer II, between 5.8S and 28S of the rRNA gene, was amplified with the 3' primer ITS4. PCR reactions were carried out with 50ng genomic DNA in MyTaq Red mix (Bioline (Aust) Pty Ltd). PCR conditions were: initial denaturation step at 94°C for 2 minutes, followed by 35 cycles of 1 minute at 94°C, annealing at 55°C and extension for 1.5 minutes at 72°C and a final elongation step of 72°C for 2 min (Gunasinghe et al., 2016a). PCR products were separated and visualized by gel electrophoresis at 84V for 45 minutes on a 1% (w/v) agarose gel and products of 500bp (20µl of each) were sequenced by Macrogen Inc. (Korea).

RESULTS AND DISCUSSION
Observations confirmed the morphogenic plasticity of N. capsellae. The species changed its morphology from septate hyphal growth to budding yeast form or vice-versa depending on external conditions. The morphological transformation (hyphae-to-yeast) of N. capsellae resulted in two common single-celled blastospore (blastoconidia) types distinguished based on gross appearance: The first, predominantly elliptical and comparatively large single-celled meso-blastospores that resembled yeast cells where average dimensions were 4.8 x 2.7µm (Table 1 and Figure 1A) and the second, single-celled micro-blastospores (average dimensions of 1.9 X 0.9 µm) that resembled rod-shaped bacterial cells (Table 1 and Figure 1B) and comparatively large, spherical to cylindrical arthrospores (arthroconidia) ( Figure 1C). The formation of blastospores (Odds and Kerridge, 1985;Gauthier, 2015;Lin et al., 2015) and arthrospores (Barrera and Szaniszlo, 1985) are important aspects of fungal morphogenesis in many dimorphic or polymorphic fungal species, including Ascomycetes. Blastospores are asexually produced single-celled spores that can increase in number rapidly through repeated budding (Odds and;Kerridge, 1985;Mehrabi et al., 2006;Lin et al., 2015), while arthrospores are formed by segmentation and subsequent fragmentation of existing hyphae (Barrera and Szaniszlo, 1985). This morphological transition in N. capsellae has not been previously recognized. To eliminate the potential for contamination being the source of the newly described structures, the identity of the structures was confirmed by PCR amplification and sequencing of the ITS1 and ITS4 regions of nuclear ribosomal DNA of blastospores taken from a yeast-like colony (GenBank Acc. No. MZ149251) and hyphae from a pure mycelial colony (GenBank Acc. No. MW898135). The BLAST results for both samples were 99% match to N. capsellae.

Pure Cultures on Growth Media
The formation of single-celled spores (blastospores and arthrospores) produced prominent changes in colony morphology (Figures 2 and 3). Morphological transformation from hyphae-to-yeast or yeast-to-hyphae was observed on all 5 tested media, although some different media favoured the formation of yeast and the other the formation of mycelial phase. Transferring hyphae to a fresh medium (subculturing) initiate yeast-like colonies on some media demonstrating hyphae-to-yeast transformation. Generally, on commonly used media like PDA or NDY, and WA, the mycelial phase dominated with an infrequent, short, initial yeast-like phase immediately after subculturing. When yeast-like growth was observed initially (Figures 2A, B), yeast-to-hyphae transition was initiated in colonies anytime within two-three weeks ( Figures 2C, D). The hyphae-to-yeast transformation was also observed in mycelial colonies older than one month (Figure 3). At the initial growth stage on some media such as NDY, PDA, and WA, the yeast phase was present but visible only under the microscope and the proportion of yeast to mycelia and the time taken for the yeastto-hyphae transformation differed across the media tested.
Subculturing and the method of inoculation influenced the morphological transformation of the fungus. Placing N. capsellae hyphal fragments onto a new growth medium promoted the initial formation of creamy white-to-tan-coloured colonies dominated by single-celled blastospores (Figures 2A, B). Similar growth behaviour on agar media by dimorphic pathogens has been observed. Initially formed yeast-like colonies of Candida citri switched to mycelial growth after prolonged incubation (Sipiczki, 2011). Odds and Kerridge (1985) claim that the C. albicans blastospore-dominating yeast phase was easy to obtain and maintain on artificial media, while the pure mycelial phase was difficult to attain as it is only a transient form.
Of the two inoculation methods tested, streaking favoured the formation of the yeast phase. Hyphal growth with or without initial yeast-like growth was observable on PDA, NDY, and V8 juice agar when inoculated with an agar plug ( Figures 4A, B, F). Values are the mean of more than 400 cells from each morphotype measured in-vitro and in-planta. Except for WA, yeast-like colonies appeared on streak plates of all media within 24h of inoculation ( Figures 4C, D). On WA, yeast-like colonies, hyphal growth, or both were observed during the initial growth stage depending on the inoculation method used ( Figure 4E). The appearance of initial yeast-like colonies immediately after subculturing was fast, within overnight incubation ( Figures 4C, D), while it took at least 10-14 days for hyphal growth to occur when the initial yeast growth was absent ( Figures 4E, F). The co-occurrence of cells with different morphologies varying from single-celled blastospore types to true hyphae in a single colony was observed in N. capsellae colonies ( Figure 5). However, the dominating morphotype and the architecture of the colonies seem to depend on several factors, such as isolate,  method of inoculation, and type of the growth medium and changed over time with the age of the colony (Figures 2 and 3).
The formation of mixed colonies with different cell morphologies has been reported in other dimorphic pathogens such as Cryptococcus (Levy et al., 2012;Lin et al., 2015) and Candida sp. (Odds and Kerridge, 1985;Sipiczki, 2011). Cryptococcus produces colonies with different morphologies, where yeast cells (blastospores) predominate in the colony centre and hyphae dominate at the periphery (Lin et al., 2015). Obtaining a colony consist of single-cell morphology was a difficult task for C. albicans on artificial culture media because of the co-existence of blastospores, hyphae, and pseudohyphae (Odds and Kerridge, 1985). Such colonies, the architecture of the colony, degree of morphological heterogeneity, and the proportion of each cell morphology of otherwise genetically identical cells, are a result of differential gene expression among cells (Levy et al., 2012) as influenced by one or more of several factors, such as gradient of signalling molecules created by the varying microenvironments across the growth medium and/or factors like the age or cellcycle stage (Lin et al., 2015). Apart from much larger colonies at 37°C, the same growth patterns were observed at both temperatures (23°C and 37°C). Therefore, these in-vitro studies at two temperatures suggest that N. capsellae morphogenesis is not induced solely by temperature as seen in thermally dimorphic human fungal pathogens (Dukik et al., 2017), but by more complex external and internal factors, including the growth medium.

Multi-Celled Macroconidia in Sterile Distilled Water
Multi-celled macroconidia produced both micro-blastospores and meso-blastospores or arthrospores in-vitro within a 24h incubation period. All three spore types were stained with cotton blue and observable under LM among the multi-celled macroconidia when suspended in sterile distilled water ( Figure 6).
However, not all multi-celled conidia produced single-celled blastospores, as some germinated by producing single or multiple germ tubes.

In-Planta Morphogenesis
Multi-celled macroconidia or hyphal fragments newly introduced onto host tissue (cotyledons) formed single-celled blastospores or arthrospores displaying the same trend of morphological transition that occurred in-vitro. All singlecelled spore types similar to the spore types produced in-vitro (meso and micro-blastospores and arthrospores) were evident on cotyledons inoculated with multi-celled macroconidia  (Table 1 and Figures 7A, B, E, F) or mycelial fragments ( Figures 7C, D) within 48h post-inoculation and this was consistent across all six test genotypes belonging to three different host species. However, germinated multi-celled macroconidia with single or multiple germ tubes were also common. Initial observations were that the amount of germinating multi-celled macroconidia or each morphologically different single-celled blastospore type could vary depending on the host, such as susceptible and resistant host species. After landing on the host tissue, the formation of a single-celled blastospore (yeast phase) on the host surface just before invasion has been observed for several dimorphic plant pathogens (Ingold, 1995;Ruiz-Herrera et al., 1995;Vollmeister et al., 2012;Cisséet al., 2013). Asexually produced diploid, multi-celled pycnidiospores of Zymoseptoria tritici, a dimorphic wheat pathogen causing Septoria tritici blotch (Palmer and Skinner, 2002;Mehrabi et al., 2006) or teliospores of Ustilago maydis, a maize pathogen causing smut disease (Ruiz-Herrera et al., 1995;Vollmeister et al., 2012), establish a short yeastlike phase on the leaf surface by producing blastospores before host invasion in the filamentous form. In Taphrina deformans, the cause of leaf curl disease of peach and nectarine, ascospores discharged from curled leaves produce single-celled blastospores, that overwinter and then switch back to the infectious filamentous form to invade the host (Rodrigues and Fonseca, 2003;Tavares et al., 2004;Cisséet al., 2013). This morphological transition from multi-celled conidia or hyphae-to-yeast and yeast-to-hyphae plays unique species-specific roles in virulence (Mehrabi et al., 2006;Gauthier, 2015;Lin et al., 2015). For instance, yeast cells of T. deformans function as an overwintering dormant spore (Cisseé t al., 2013), while dissemination is the key role of yeast cells produced by Ophiostoma ulmi (Lin et al., 2015).
Although there can be a variety of different morphologies observed, the formation of blastospores involves three steps of bud emergence, bud growth, and spore/conidia separation. The sites at which blastospores are produced during hyphae-to-yeast transformation could be limited and for C. albicans these hyphal sites are adjacent and posterior to septa (Odds and Kerridge, 1985). Similarly, in N. capsellae yeast-like meso-blastospore formation occurred at the septa of hyphae ( Figures 8A, B) or the tip ( Figure 8B). However, micro-blastospore formation from hyphae was random and occurred in comparatively large numbers ( Figures 8C, D).

Arthrospore (Arthroconidia) Formation
The formation of single-celled arthrospores from existing hyphae has long been observed during fungal morphogenesis in certain fungal species (Miyazi and Nishimura, 1971;Bibel et al., 1977;Emyanitoff and Hashimoto, 1979;Cartwright et al., 2014). These single-celled spores are produced in extensive numbers to enhance the pathogen's reproductive potential and, therefore, the primary function of arthrospore is reproduction rather than survival between seasons. However, arthrospores can also be the principal means of dissemination in certain pathogenic fungi (Barrera and Szaniszlo, 1985). Arthroconidia formation and morphogenesis in dimorphic human fungal pathogens can be stimulated by several types of environmental stress conditions related to growth medium composition, carbon dioxide tension, and increased temperatures (Rippon and Scherr, 1959;King and Jong, 1976;Bibel et al., 1977;Emyanitoff and Hashimoto, 1979). Initiation of arthro sporogenisis, the age of conidia formation, and the number of conidia formed are all highly species-specific (Barrera and Szaniszlo, 1985). Arthro sporogenesis of N. capsellae was observed in-vitro from multi-celled hyphae on different artificial media (Figure 9) or multicelled macroconidia (Figure 10). This is the first report showing arthrospores from N. capsellae, although the cleavage of multi-celled macroconidia on cotyledons to produce single cylindrical entities by this pathogen has been reported earlier by Gunasinghe et al. (2016d). Two different morphologies commonly observed for N. capsellae arthrospores were cylindrical ( Figures 9A, B) or spherical ( Figures 9C, E). Initial observations are that the size of the arthrospores and the proportion of each morphology in-vitro vary depending on the growth conditions. This is not unusual as other dimorphic fungal species can produce a range of arthroconidia that differ in size and/or shape and the dominating type is determined by the substrate (King and Jong, 1976;Allermann et al., 1978;Kier et al., 1980;Barrera and Szaniszlo, 1985).
The development and release of arthroconidia appear to vary depending on the fungal species (Barrera and Szaniszlo, 1985). In both: Mucor rouxii and Tricophyton mentagrophytes, arthroconidia develop inside the hyphae as cylindrical conidial chains enveloped by the original hyphal cell wall but then change their morphology to spherical just before dehiscence (Takeo and Nishiura, 1974;King and Jong, 1976). Separation of the conidia could occur by autolysis of inter-septal material in the original hyphal wall that helps the adhesion of neighbouring septal walls of conidial chains or by dehiscence of the original hyphal wall (Barrera and Szaniszlo, 1985). The arthroconidia of Antromycopsis brousonnetiae developed a secondary cell wall inside the hyphae and the separation of conidia is by centripetal dehiscence of the original hyphal wall (Moore, 1977).
The frequently observed release mechanism for N. capsellae arthrospores produced in-vitro was dehiscence of original hyphae. Transverse dehiscence at the septa was more common (Figures 9C-E) than centripetal dehiscence ( Figure 9F). However, the release of cylindrical arthrospores by autolysis of the original hyphal wall also occurred when arthrospores were produced from multi-celled macroconidia ( Figure 10). There is evidence that arthrospores are a superior asexual spore type providing enhanced dissemination and better establishment on the host tissue by protecting cells from host lytic enzymes  (Duran et al., 1973), increasing antibiotic production (Nash and Huber, 1971;Queener and Ellis, 1975), and by promoting surface adherence (Kurakado et al., 2020).
Morphological Plasticity, Pathogen Biology, and Disease Epidemiology FIGURE 10 | Arthrospores produced by Neopseudocercosporella capsellae multi-celled macroconidia on the host. (A) Cylindrical arthrospore formation by disintegration of original macroconidia cell walls at septa (arrows) on the cotyledon surface as seen under the light microscope. (B) Scanning electron micrograph of multi-celled macroconidia (c) and germinating three arthrospores as single-celled spores by producing a single germ tube each (arrows) on the cotyledon. Scale bars = 10µm.  stimuli. During its morphogenesis, dikaryotic mycelium or asexually produced multi-celled macroconidia (2n) on leaf lesions in their parasitic phase initiate a single-celled yeast phase through the production of at least three types of morphologically distinguishable single-celled spore types. Multicelled macroconidia on the host surface develop in three possible ways 1) they separate from the septa to produce multiple arthrospores that germinate as single entities, 2) they produce blastospores that increase in number rapidly by budding, and 3) they germinate as typical conidia by producing single or multiple germ tubes. This morphological plasticity of N. capsellae can be considered a cost-efficient strategy to provide a more resilient form of spores on the host enhancing inoculum potential and dissemination, both contributing to the pathogen's virulence. A systematic study is recommended to fully understand N. capsellae multifaceted life cycle and to define the role of morphogenesis in relation to virulence, which is largely unknown.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

AUTHOR CONTRIBUTIONS
NG: Designed and conducted the work and drafted the manuscript. PD: Worked with NG to conduct the study and prepared the figures. SN, MB, and MY: Gave significant contribution to the conception and the design and reviewed the manuscript for critical and important intellectual content. All authors contributed to the article and approved the submitted version.

FUNDING
Funding support for this research was provided by the School of Plant Biology at the University of Western Australia, and the Centre of Crop Health at the University of Southern Queensland.