Development of a transformation system for Hirsutella spp. and visualization of the mode of nematode infection by GFP-labeled H. minnesotensis

Hirsutella rhossiliensis and H. minnesotensis are endoparasitic fungi of the second-stage juvenile (J2) of the soybean cyst nematode (Heterodera glycines) in nature. They also parasitize both H. glycines J2 and Caenorhabditis elegans on agar plates. Agrobacterium tumefaciens-mediated transformation conditions were established for these Hirsutella spp. The resulting transformants were similar to the corresponding wild-type strains. The infection processes of H. glycines J2 and C. elegans second larval stage (L2) by H. minnesotensis expressing ZsGreen were microscopically analyzed. Conidia of H. minnesotensis adhered to passing nematodes within 8 h post-inoculation (hpi), formed an infection peg between 8 and 12 hpi, and penetrated the nematode cuticle between 12 and 24 hpi for C. elegans L2 and between 12 and 32 hpi for H. glycines J2. Hyphal proliferation inside of the nematode coelom was observed at approximately 32 hpi for C. elegans L2 and at approximately 40 hpi for H. glycines J2. The fungus consumed the whole body and grew out to produce conidia at approximately 156 and 204 hpi for C. elegans L2 and H. glycines J2, respectively. The efficient transformation protocol and a better understanding of infection process provide a solid foundation for studying the molecular and cellular mechanisms underlying fungal parasitism of nematodes.


Results
Transformation of Hirsutella spp. with three fluorescent protein genes. The sensitivities of Hirsutella spp. to hygromycin B and geneticin were determined. The minimum concentrations of hygromycin B and geneticin resulting in complete growth inhibition were 150 μ g/ml and 400 μ g/ml, respectively, for H. minnesotensis AS3.9869 and 100 μ g/ml and 400 μ g/ml, respectively, for H. rhossiliensis AS6.0004. Growth responses of these isolates to hygromycin B on potato dextrose agar (PDA) and cornmeal agar (CMA) were indistinguishable. Based on their higher sensitivity to hygromycin B than geneticin, we chose the hph gene, which confers resistance to hygromycin B, to select transformants. Binary vectors containing hph and the AsRed, AsCyan or ZsGreen gene were used for transformation. PCR analysis of 12 randomly selected transformants revealed that the hph gene in the inserted T-DNA was detected in all selected transformants but not in the wild-type (wt) AS3.9869 or AS6.0004 strains (Fig. 1). Microscopic observations of slide cultures showed that all three fluorescent proteins were expressed strongly in hyphae and conidia of AS3.9869 and AS6.0004 transformants (Fig. 2). Morphology, growth characteristics and the capability to parasitize nematodes were evaluated for 50 transformants (10 ZsGreen and 10 AsCyan transformants of AS3.9869; and 10 ZsGreen, 10 AsCyan and 10 AsRed transformants of AS6.0004). Compared with the corresponding wt strains, several transformants parasitized fewer nematodes, which may be due to genetic changes that occurred during transformation. However, most transformants showed no significant differences from the wt strains.    16, and 24 h post inoculation (hpi) showed that even at 4 hpi, more than 90% of H. glycines J2s had attached conidia (Fig. 4). The percentage of C. elegans L2s with attached conidia at this stage was approximately 65%, which was significantly lower than that observed for the H. glycines J2s. The percentage of parasitized H. glycines J2s did not vary greatly during this period (Fig. 4), suggesting that unattached H. minnesotensis conidia quickly recognized and adhered to passing H. glycines J2s. There was a noticeable increase in the colonization of C. elegans L2s between 4 and 8 hpi, after which it remained stable at approximately 75%. More conidia adhered to individual nematodes as time progressed, with up to 20 adhered spores for each nematode until mobility was lost. These results suggest that the stage of recognition and adhesion was mostly completed within 8 hpi for both nematodes (Fig. 5-7).
Penetration. Following the germination of the conidia adhered to the nematodes, an infection peg was formed, and penetration of the nematode cuticle occurred between 8 and 12 hpi (Figs. 5-7). After penetration, the apex of the infection hypha became globus between 12 and 24 hpi for C. elegans L2 (Fig. 5C, Fig. 7) and between 12 and 32 hpi for H. glycines J2 (Fig. 6C, Fig. 7). Although C. elegans L2s were alive and could move slowly at 24 hpi, more than 80% of them were observed to be penetrated by one or more conidia. H. glycines J2s appeared dead at 32 hpi because their stylets were no longer moving, and more than 90% of them were penetrated by one or more conidia during this period (Fig. 7).

Consumption of the nematode body.
Once the fungus penetrated the nematode, the initial hypha started to grow in the nematode body. The fungus began to develop an initial hypha inside of C. elegans L2 (Fig. 5D) at 32 hpi and inside of H. glycines J2 at 40 hpi (Fig. 6D). Eighty percent of the examined nematodes with developed initial hyphae were found at 48 hpi for C. elegans L2 and at 60 hpi for H. glycines J2. The initial hyphae quickly extended inside of C. elegans L2 (

Discussion
Green fluorescent protein (GFP) and its color variants as molecular labels have been applied to studying diverse filamentous fungi 28 . To assess the interactions between nematophagous fungi and nematodes, attempts have been made to transform Pochonia chlamydosporia 23 , Clonostachys rosea 26 and Dactylellina cionopaga 29 with a gene encoding GFP. However, no gfp transformant of P. chlamydosporia was obtained 23 because of its slower growth and limited sporulation, and the material coating its spores resulted in a very low transformation efficiency 26 . The goal of our study was to establish a robust transformation protocol for the dominant nematode endoparasitic fungi Hirsutella spp. in nature using fluorescent protein genes as markers to assess the manner by which they infect nematodes. ATMT was successful in transforming both H. rhossiliensis and H. minnesotensis, and most transformants expressed three fluorescent proteins with high efficiency and displayed growth, morphology, and pathogenicity that were indistinguishable from the corresponding wt strains. These transformants will facilitate ecological studies, such as the monitoring of the spatial and temporal distribution of Hirsutella spp. in soil and plant roots, competition Due to the movement of free-living nematodes, their microscopic observation is challenging. Conventional techniques for immobilizing nematodes include the uses of cyanoacrylate glue 30 and anesthetic compounds, such as sodium azide (a metabolic inhibitor) 31 and levamisole (a cholinergic agonist) 32 . Recently, a computer-controlled micro-fluidic device was developed to limit nematode mobility 33 . Although these methods have been successfully applied, some limitations remain. The use of the glue is labor-intensive and low-throughput. Anesthetic compounds have negative effects on H. minnesotensis growth. Further, microfluidic device cannot provide enough air for this fungus. Here, we successfully H. minnesotensis is a cold-adapted fungus that preferentially parasitizes cyst nematodes, even in the presence of bacterial-feeding nematodes in nature 19,34 . Both H. minnesotensis and H. rhossiliensis can parasitize various nematodes, including C. elegans, on agar plates 15,22 . However, they cannot parasitize bacterial-feeding nematodes in natural soil, or C. elegans in autoclaved soil, as demonstrated in our recent study. The differences in parasitism between agar plates and soil suggest that some soil factors may interfere with fungal recognition and/or colonization of nematodes. Follow-up studies to address this supposition will be conducted using the transformants generated in this study.
Several characteristics of C. elegans, including its small size, short generation time, invariant developmental lineage, and sequenced genome, along with the powerful genetic manipulation tools available for its analysis and its ease of handling, have made it a model organism for studying the mechanisms underlying bacterial and fungal pathogenesis in animals and innate immunity 35 . Studies with bacterial pathogens, such as Pseudomonas, Salmonella, or Serratia, have been highly fruitful in revealing the existence of an innate immune system in C. elegans, similar to that in vertebrates 36 . This model organism has also been used to study innate immune responses to fungal infections 37,38 . Interactions between nematophagous fungi and nematodes have not been extensively studied.
The capability of H. minnesotensis, a dominant H. glycines J2 parasite, to infect C. elegans on agar plates presents a new model system for investigating the molecular and cellular mechanisms underlying the pathogenesis of nematophagous fungi and host defenses at different infection stages. The infection process of H. minnesotensis against H. glycines J2 and C. elegans L2 observed in this study has guided transcriptomics analyses of C. elegans L2 parasitized by H. minnesotensis 39 , as well as an ongoing study of the experimental evolution of parasitism of Hirsutella spp. Three A. tumefaciens strains, including SK2245 (A. tumefaciens strain EHA105 that carries a pBHt2 binary vector containing the AsRed gene), SK2247 (EHA105 that carries a pBHt2 binary vector containing the AsCyan gene) and SK2251 (EHA105 that carries a pBHt2 binary vector containing the ZsGreen gene), were used for fungal transformation. The expression of all three fluorescent protein genes was driven by the promoter of the Fusarium verticillioides elongation factor 1α gene. These A. tumefaciens strains were cultured at 28 °C for 48 h in minimal medium 40 supplemented with 50 μ g/ml kanamycin before transformation.

Materials and Methods
Soybean cyst nematode race 4 was isolated from a soybean field near Beijing. For H. glycines egg preparation, newly formed cysts in soybean roots cultivated in the greenhouse were extracted using the sucrose floatation and centrifugation method 18 . Individuals at the J2 stage were prepared as described by Liu and Chen 1 and used immediately. Bristol strain N2 of C. elegans in mixed-stages was cultured on nematode growth medium (NGM) plates and seeded with Escherichia coli stain OP50 41 at 24 °C. Fresh C. elegans specimens at the four larval stages were harvested following the method described by Xie et al. 42 and used immediately.

Preparation of fungal spores. Spores of H. minnesotensis AS3.9869 and H. rhossiliensis AS6.0004
were harvested by the scraping of 15-day-old cultures on PDA after the flooding of plates (5 plates per strain) with 15 ml of sterile water. After passing individual spore suspensions through two layers of Miracloth (Calbiochem, Cat. #475855, La Jolla, CA) to remove large debris, they were centrifuged at 10,000 rpm for 10 min and washed twice with sterile distilled water. The harvested spores were re-suspended in sterile water to produce a solution of 1 × 10 6 spores/ml.

Fungal transformation.
To determine the sensitivities of the fungal strains to antibiotics prior to transformation, PDA and CMA plates with different amounts of hygromycin (50, 100, 150, 200, 250 μ g/ ml) or geneticin (50, 100, 200, 400, 800 μ g/ml) and control plates without antibiotics were prepared. Small amounts of mycelia from the tested strains were inoculated on each plate and incubated for 15 days. The amounts of antibiotics that completely inhibited fungal growth were chosen for transformant selection. Agrobacterium tumefaciens-mediated transformation was performed according to Khang et al. 40 with minor modifications. Each of the three A. tumefaciens strains described above was incubated in 1 ml of minimal medium (MM) supplemented with kanamycin (50 μ g/ml) for 1-2 days at 28 °C. The resulting A. tumefaciens cultures were inoculated into induction medium (IM) at OD 600 = 0.15. After the culturing of A. tumefaciens cells for 6 h at 28 °C at 200 rpm, a 100-μ l suspension of A. tumefaciens cells and 100 μ l of fungal spores (1 × 10 6 spores/ml) were mixed and spread onto a nitrocellulose membrane (Whatman Cat. #7141 104; 47-mm diameter; 0.45-μ m pore size) placed on co-cultivation medium in the presence of 200 μ M AS. After 2 days of incubation at room temperature in the dark, the membrane was transferred to CMA selection medium containing hygromycin B (150 μ g/ml for AS3.9869 and 100 μ g/ml for AS6.0004) and cefotaxime (0.2 mM). The membrane was incubated for 2-3 weeks at room temperature until colonies formed.
Putative hygromycin B-resistant colonies were transferred to 24-well plates containing CMA selection medium with hygromycin B using sterile toothpicks. After the plates were incubated at 25 °C for 10 days, individual wells were flooded with sterile water. The water was pipetted up and down several times to dislodge the conidia. The conidia were spread onto water agar plates with hygromycin B and incubated for 24 h, and 10 single germinating spores for each transformation were then picked up with a sterile needle under a microscope and transferred to PDA plates for subsequent analysis and preservation.
Characterization of putative transformants. A simple slide culture technique was used for detecting the fluorescence of the growing transformants. A PDA agar block (approximately 4 × 4 mm 2 and 4 mm thick) inoculated with fungus was placed on a sterile slide with a coverslip 43 . The slide culture was placed into a 90-mm Petri dish sealed with parafilm to avoid contamination and to maintain humidity. After 7-10 days of incubation at room temperature, the mycelia were observed with a Zeiss Axio Imager A1 fluorescent microscope (Carl Zeiss, Germany) with AxioVision 4.6 software to confirm the expression of each of the three fluorescent proteins. Three filter sets were used as follows: (i) excitation at 460-480 nm and emission at 505-530 nm for ZsGreen; (ii) excitation at 540-552 nm and emission at 575-640 nm for AsRed; and (iii) excitation at 365 nm and emission at 420-470 nm for AsCyan.
Twelve randomly selected fluorescent transformants were cultured on cellophane membranes placed on PDA plates containing hygromycin B at 25 °C for 10 days. Mycelia were harvested from the cellophane, frozen in liquid nitrogen and ground to a fine powder. Fungal DNA was extracted using the CTAB method and purified using a Wizard® Genomic DNA Purification Kit (Promega, USA) according to the manufacturer's instruction. The primers hph-F (5'-GAGCCTGACCTATTGCATCTC) and hph-R (5'-CCGTCAACCAAGCTCTGATAG) were used for PCR analysis to confirm the presence of the integrated hph gene following the protocol recommended by White and Chen 44 , with an initial denaturation step at 95 °C for 5 min, 35 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and elongation at 72 °C for 30 s and a final elongation step at 72 °C for 10 min. PCR products were separated in a 1% agarose gel in TAE buffer. The D2000 DNA ladder (TIANGEN Biotech, China) was used as a size marker.
Evaluation of transformant stability and capacity to parasitize nematodes. Single-spore purified transformants were cultured on PDA plates without hygromycin B for at least five generations, and they were then evaluated for the stable production of individual fluorescent proteins.
To assess the pathogenicity of the selected transformants, wt strains and their transformants were inoculated on PDA plates at 25 °C. Individual conidial suspensions were prepared by washing 2-week-old colonies with sterilized 0.1% Tween-20 solution. The number of conidia was determined using a hemacytometer, and the conidial concentration was adjusted to 2× 10 5 spores/ml. An aliquot of the conidial suspension (200 μ l) was spread onto the surface of each CMA plate. The plates were left uncovered for 30 min in a laminar flow hood to allow excess water to evaporate. After 2 weeks of incubation at room temperature, approximately 300 nematodes in 100 μ l of 4.5 mM KCl solution were added to the center of each plate 45 . After 3 days of incubation, the nematodes were washed off using 5 ml of 0.1% Tween-20 solution. The percentage of nematodes with attached conidia or of those colonized by each strain was Scientific RepoRts | 5:10477 | DOi: 10.1038/srep10477 determined by observing 100 randomly chosen nematodes from each plate using an Olympus CK 40 microscope (Olympus, Japan) at 40-200× magnifications 18 . Four plates were used for each strain.

Microscopic observation of infection process.
To determine the effect of inoculation time on the percentage of parasitism by H. minnesotensis, 200 H. glycines J2s or C. elegans L2s were inoculated in each well of a 12-well plate containing 2-week-old fungal colonies growing on CMA at 25 °C. The nematodes were washed out with 1 ml of sterile 0.1% Tween-20 at different time points (4,8,16, and 24 h). The nematodes that were parasitized or possessed adhered conidia were counted.
After 4 h of incubation, the nematodes from each well of the 12-well tissue culture plates were collected using sterile water and transferred into each well of a 6-well tissue culture plate to observe parasitism and to select nematodes with a single conidium attached. One drop of 4% boiled agarose was placed on the center of a sterilized slide, and a sterilized coverslip was placed onto the agarose drop immediately and pressed down to spread the agarose 46 . Once the agarose was solidified, the coverslip was removed. One nematode with a single spore attached was picked up and transferred to the center of the agarose pad and covered with a coverslip. Five slides were prepared for each nematode species. The location of the nematode was marked, and three edges of the coverslip were sealed with Vaseline. Another 30 nematodes with single or multiple spores attached were picked up in the same manner and transferred to 5 different agarose pads as described above. The slides were kept in a container at a relative humidity of 98% (saturated solution of K 2 SO4) 47 at 20 °C. The infection process was observed and photographed with a Zeiss fluorescence microscope (Axio Imager A1) with Apochromat 20× and 40× objective lenses, and AxioVision 4.6 software was used for image acquisition.

Statistical analysis.
The effects of the nematode species, stages and fungal species and isolates on the percentages of parasitized nematodes were determined using analysis of variance (ANOVA). The significances of the differences between the percentages of nematodes parasitized by AS3.9869 and AS3G1 were analyzed using the t-test. The SPSS (Statistical Package for the Social Sciences) software package (SPSS 16.0) was used for all statistical analyses. The data pertaining to the infection cycles of H. minnesotensis in C. elegans L2 and H. glycines J2 were analyzed using the MeV software package (MeV 4.8.1).