**1. Introduction**

The filamentous ascomycete *Metarhizium robertsii* is an omnipresent and soil-dwelling pathogen of insects, ticks and mites [1,2]. It has been developed as a promising mycoinsecticide to control different insect pests and investigated as a genetically tractable system for studying fungus–insect interactions [3–5]. Similar to plant pathogens like the rice blast fungus *Magnaporthe oryzae*, *M. robertsii* infects insect hosts by penetrating host cuticles through the differentiation of the infection structure appressoria and build-up of the turgor pressure within appressorial cells [6,7]. The generation of the turgor pressure requires the accumulation of high concentrations of glycerol and or solutes within appressorial cells [8–10], which are separated from conidial mother cells by the formation of septa with the central septum pores sealed [11,12]. The mechanism of septal pore sealing in insect pathogens like *M. robertsii* is still unclear.

Fungal hyphal cells are separated by perforate septa, and filamentous fungi evolved with finely tuned strategies to balance the inter-cellular exchanges and the need for compartmentalization [13]. For ascomycete fungi belonging to the Pezizomycotina subphylum, Woronin bodies (WBs) are formed for plugging/unplugging the septal pores to regulate organelle exchanges between compartments, maintain hyphal cell heterogeneity and prevent excessive cytoplasmic bleeding in the event of hyphal damage [13–15]. WB is a peroxisome-type and hexagonal crystal-like organelle, which is membrane-bound and contains a dense core developed from the self-assembly of a single protein Hex-1, which has been first characterized in the model fungus *Neurospora crassa* in a very close association with septa [16,17]. The Hex-1-like proteins (either called HexA or Hex1) have then been identified and characterized in a few fungal species such as *Aspergillus oryzae* [15], *A. fumigatus* [18] and the plant pathogenic fungi like *M. oryzae* [12] and *Fusarium graminearum* [19]. The deletion of the *Hex1* gene resulted in the disappearance of WBs in fungal hyphae and the null mutants demonstrated impaired stress resistance abilities against the osmotic and cell-wall integrity interference agents, a dramatically reduced ability to survive wounding and or a reduced capacity in the infection of hosts [18,19]. For example, the *Hex-1* null mutant of *N. crassa* had reduced growth on a minimal medium and was impaired in sporulation [20]. After the deletion of *Aohex1* in *A. oryzae*, septal plugging was abolished and hyphal heterogeneity also a ffected [15]. The HexA of *A. fumigatus* was verified to be important for stress resistance and virulence [18]. For plant pathogenic fungi, it has been revealed that the formation of the Hex1-associated WB was required in *M. oryzae* for the development and function of the infection structures appressoria and therefore host colonization [12]. The disruption of the *Hex1* gene in *F. graminearum* reduced fungal asexual reproduction, infectivity and virus RNA accumulation in the infected cells when compared with the wild-type strain [19]. Likewise, the homologous *Hex1* gene was found to be required for WB formation, conidiation and the formation of the capturing trap in the nematophagous fungus *Arthrobotrys oligospora* [21]. The gene(s) responsible for WB formation and function in ascomycete entomopathogenic fungi has ye<sup>t</sup> to be investigated.

In this study, it is intriguing to find the substantial length variation amongs<sup>t</sup> the Hex1-domain-containing proteins from di fferent fungi. We then performed the loss-of-function investigation of a homologous *Hex1* gene (MAA\_00782, designated as *Mrhex1*) in the insect pathogenic fungus *M. robertsii*. It was found that *Mrhex1* was required in *M. robertsii* for WB formation, asexual growth and sporulation, appressorium di fferentiation and the topical infection of insect hosts. In contrast to the findings in other fungi, however, the null mutant of *MrHex1* could tolerate di fferent stress conditions like the wild-type strain.

#### **2. Materials and Methods**

#### *2.1. Strains and Culture Conditions*

The wild-type (WT) strain and mutants of *M. robertsii* strain ARSEF 2575 were routinely cultured on potato dextrose agar (PDA; BD Difco, Franklin Lakes, USA) at 25 ◦C. Spore germination and appressorium induction assays were conducted using locust (*Locusta migratoria manilensis*) hind wings or the minimal medium (MM) (NaNO3,6g/L; KCl, 0.52 g/L; MgSO4·7H2O, 0.52 g/L; KH2PO4, 0.25 g/L) amended with 1% glycerol as the sole carbon resource (MM-Gly) [22]. For genomic DNA, RNA extractions and hyphae staining, fungal spores were cultured in Sabouraud dextrose broth (SDB; BD Difco, Franklin Lakes, USA) for three days at 25 ◦C and incubated at 200 rpm in a rotary shaker.

#### *2.2. Protein Feature Characterization and Phylogenetic Analysis*

Homologous Hex1 proteins were retrieved from GenBank for those containing the conserved S1\_Hex1 domain (Table S1). The conservation analysis of the S1\_Hex1 domains of 460 proteins obtained from di fferent fungal species/strains was characterized with the program WebLogo (ver. 2.8.2) [23]. For phylogenetic analysis, 21 proteins selected from representative fungal species were aligned with the program Clustal X ver. 2.0 [24], and a bootstrapped (1000 replicates) neighbor-joining (NJ) tree was constructed with the program MEGA X [25] using the pairwise deletion of the alignment gaps and a Dayho ff substitution model.

#### *2.3. Gene Deletion and Complementation*

To determine the function of *MrHex1*, targeted deletion was performed by homologous recombination via the *Agrobacterium*-mediated transformation of the WT strain of *M. robertsii* as described before [26]. In brief, the 5- and 3- flanking sequences were amplified using the genomic DNA as a template with the primer pairs hex1UF (CGGAATTCGTACGGACCGATAAAACGTG) and hex1UR (CGGAATTCGAATGTCCTCCTTGATGTC), hex1DF (GCTCTAGACTGTCGACTGC-TTTCGAGTC) and hex1DR (GCTCTAGATAAGACACCCCATGTCAGC), respectively. The products were digested with the restriction enzymes *EcoR*I and *Xba*I, and then inserted into the same enzyme-treated binary vector pDHt-bar (conferring resistance against ammonium glufosinate) to produce the plasmid pBarhex1-KO for fungal transformation. For null mutant complementation, the full open reading frame (ORF) of the *Mrhex1* gene was amplified together with its promoter and terminator regions using the primer pairs hex1U (GGACTAGTGCACAGAGGACAAAACATGG) and hex1L (GGACTAGTTTACAGGCGAGAGCCGTGAA). The product was digested with *Spe*I, and then inserted into the binary vector pDHt-ben to produce the plasmid pBenhex1 (conferring resistance against benomyl) [27]. The drug-resistant mutants were isolated and verified by PCR and reverse transcription-PCR (RT-PCR) analyses with the primers hex1F (CACCACCACCATGACCAC) and hex1R (GAGAGCCGTGAATGACCTT). The β-tubulin gene (MAA\_02081) was used as the control and amplified using the primers TubF and TubR [28].

#### *2.4. Phenotyping, Cell Integrity and Stress Response Assays*

To determine the e ffect of *MrHex1* deletion on fungal growth and conidiation, fungal cultures were grown on PDA and the colony diameters were measured at di fferent times post inoculation. After growth for 18 days, conidial production was assayed and compared between the WT and mutants by two-tailed Student's *t*-tests [29]. To determine cell integrity after gene deletion, the level of cellular content leakage was determined via the detection of free amino acids in liquid culture filtrates by reaction with ninhydrin [30]. Thus, the spores of the WT and mutants were collected from 14 day-old PDA plates and inoculated in SDB at 25 ◦C and 200 rpm for 4 days. Fungal mycelia were collected by filtration, washed twice with sterile distilled water and transferred to MM–N (i.e., without the addition of NaNO3 in MM) liquid medium for 24 h. The supernatants were collected by filtration and transferred (4 mL each) into the test tubes followed by the individual addition of 1 mL of 2% ( *w*/*v*) of ninhydrin reagen<sup>t</sup> and 1 mL phosphate bu ffer (pH, 8.0). The samples were mixed by vortexing prior to being treated in a boiling water bath for 15 min. After cooling at room temperature, the absorbance of each sample was recorded at 570 nm (A570) using a Biophotometer (Eppendorf). The reaction solutions were also transferred into 1.5 mL centrifuge tubes for photographing. The corresponding mycelia of each strain were dried in an oven at 50 ◦C overnight and weighed. The unit of A570 was then normalized with the mycelium dry weight of each sample. There were three replicates for each strain and the experiments were repeated twice. The two-tailed Student's *t*-tests were conducted to compare the di fferences between strains.

For stress challenges, fungi were grown on PDA or PDA amended with the final concentrations of 0.01% sodium dodecyl sulphate (SDS), 200 μg/mL Calcofluor white and 250 μg/mL Congo red for cell wall integrity challenges; 50 μM farnesol for antifungal resistance, and 1.5 M KCl and 1 M Sorbitol for osmotic challenges [18,26], respectively. For inoculation, 2 μL of the 10-fold diluted spore suspensions (2 × 10<sup>7</sup> conidia/mL) were spotted onto the plates and incubated at 25 ◦C for three days.
