Divergent roles of Rad4 and Rad23 homologs in Metarhizium robertsii's resistance to solar ultraviolet damage

Abstract

The anti-ultraviolet (UV) role of a Rad4-Rad23-Rad33 complex in budding yeast relies on nucleotide excision repair (NER), which is mechanistically distinct from photorepair of DNA lesions generated under solar UV irradiation but remains poorly known in filamentous fungi. Here, two nucleus-specific Rad4 paralogs (Rad4A and Rad4B) and nucleocytoplasmic shuttling Rad23 ortholog are functionally characterized by multiple analyses of their null mutants in Metarhizium robertsii, an entomopathogenic fungus lacking Rad33. Rad4A was proven to interact with Rad23 and contribute significantly more to conidial UVB resistance (90%) than Rad23 (65%). Despite no other biological function, Rad4A exhibited a very high activity in photoreactivation of UVB-impaired/inactivated conidia by 5-h light exposure due to its interaction with Rad10, an anti-UV protein clarified previously to have acquired a similar photoreactivation activity through its interaction with a photolyase in M. robertsii. The NER activity of Rad4A or Rad23 was revealed by lower reactivation rates of moderately impaired conidia after 24-h dark incubation but hardly observable at the end of 12-h dark incubation, suggesting an infeasibility of its NER activity in the field where nighttime is too short. Aside from a remarkable contribution to conidial UVB resistance, Rad23 had pleiotropic effect in radial growth, aerial conidiation, antioxidant response, and cell wall integrity but no photoreactivation activity. However, Rad4B proved redundant in function. The high photoreactivation activity of Rad4A unveils its essentiality for M. robertsii's fitness to solar UV irradiation and is distinct from the yeast homolog's anti-UV role depending on NER. IMPORTANCE Resilience of solar ultraviolet (UV)-impaired cells is crucial for the application of fungal insecticides based on formulated conidia. Anti-UV roles of Rad4, Rad23, and Rad33 rely upon nucleotide excision repair (NER) of DNA lesions in budding yeast. Among two Rad4 paralogs and Rad23 ortholog characterized in Metarhizium robertsii lacking Rad33, Rad4A contributes to conidial UVB resistance more than Rad23, which interacts with Rad4A rather than functionally redundant Rad4B. Rad4A acquires a high activity in photoreactivation of conidia severely impaired or inactivated by UVB irradiation through its interaction with Rad10, another anti-UV protein previously proven to interact with a photorepair-required photolyase. The NER activity of either Rad4A or Rad23 is seemingly extant but unfeasible under field conditions. Rad23 has pleiotropic effect in the asexual cycle in vitro but no photoreactivation activity. Therefore, the strong anti-UV role of Rad4A depends on photoreactivation, unveiling a scenario distinct from the yeast homolog's NER-reliant anti-UV role.

Keywords: anti-UV proteins; fungal insecticidal cells; photoreactivation; solar irradiation.

Fig 1

Domain architecture, transcription profiles, and subcellular localization of Rad4A, Rad4B, and Rad23 in M. robertsii. (A) Conserved domains and NLS motifs predicted from the amino acid sequences of Rad4 and Rad23 homologs in M. robertsii (Mr) and S. cerevisiae. The decimal value associated with NLS is a maximal probability predicted. (B) Relative transcript (RT) levels of rad4A, rad4B, and rad23 during the 7-day incubation of a Mr wild-type (WT) strain at the optimal regime of 25°C and 12:12 (L:D) with respect to a standard on day 2. (C) Laser scanning confocal microscopic (LSCM) images (scale: 10 µm) for subcellular localization of Rad4A-GFP and Rad4B-GFP fusion proteins expressed in the 60-h-old WT cultures (control) grown in SDBY at the optimal regime. (D) The ratios of nuclear versus cytoplasmic green fluorescence intensities (N/C-GFI) of Rad4A-GFP and Rad4B-GFP were assessed from 12 hyphal cells (*P value from Student’s t-test). (E) LSCM images (scale: 10 µm) for subcellular localization of Rad23-GFP in the 60-h-old cultures (control) and the 58.5-h-old cultures incubated 90 min in ¼SDBY after irradiated at the UVB dose of 1.0 J/cm². (F) The N/C-GFI ratios of Rad23-GFP assessed from 15 control and stressed hyphal cells [58.5-h-old cultures incubated 90 mm in ¼SDBY containing methyl methanesulfonate (MMS) (325 µg/mL), H₂O₂ (4.41 mM) and SDS (0.03 mM) or after exposure to the UVB dose]. Different uppercase letters denote significant differences (P < 0.01 in Tukey’s test). (G) LSCM images (scale: 10 µm) for migration of Rad23-GFP into vacuoles of hyphal cells incubated 12 h in Czapek-Dox broth (CDB) free of carbon (C-free) or nitrogen (N-free) source. Images 1–4 are bright, expressed (green), DAPI (E) or FM4-64 (G) stained (red) and merged views of the same field. Error bars: standard deviations (SDs) from three cDNA samples (B) or from 12 (D) or 15 hyphal cells (F).

Fig 2

Divergent roles of Rad4A, Rad4B, and Rad23 in the asexual and infection cycles of M. robertsii. (A) Diameters of fungal colonies incubated at the optimal regime of 25°C and 12:12 (L:D) for 7 days on the plates of the minimal medium CDA and CDAs amended with different carbon or nitrogen sources and with the deletion of carbon (d[C]) or nitrogen (d[N]) source or both (d[CN]). (B) Relative growth inhibition (RGI) percentages of fungal colonies incubated at 25°C for 7 days on CDA plates supplemented with indicated concentrations of different chemical stressors (SBT, sorbitol; MND, menadione; CGR, Congo red; CFW, calcofluor white; MMS, methyl methanesulfonate; CPT, camptothecin; and HU, hydroxyurea). All colonies were initiated with ~10³ conidia. (C and D) Conidial yields were assessed from the 15-day-old cultures of all strains and the cultures of Δrad23 and control strains during a 15-day incubation at the optimal regime, respectively. Cultures were initiated by spreading 100 µL of a 10⁷conidia/mL suspension per one-fourth nutritional strength of Sabouraud dextrose agar plus yeast extract (¼SDAY) plate. (E) LT₅₀ (no. of days) estimates for fungal strains against Galleria mellonella larvae via cuticle infection [inoculated by topical application (immersion) of a 10⁷ conidia/mL suspension] or cuticle-bypassing infection (intrahemocoel injection of ~500 conidia per larva). P < 0.05*, 0.01**, or 0.001*** in Tukey’s test. Error bars: SDs from three replicates.

Fig 3

Divergent roles of Rad4A, Rad4B, and Rad23 in conidial UVB resistance and photoreactivation of M. robertsii. (A) Survival trends of conidia incubated at 25°C for 24 h in the dark after exposure to UVB irradiation at the gradient doses of 0.02–0.5 J/cm². (B) LD₅₀, LD₇₅, LD₉₀, and LD₉₅ were estimated as indices of conidial UVB resistance from fitted dose-survival trends (r² ≥ 0.97, P < 0.001 in fitness F tests). (C) Germination percentages of conidia photoreactivated for 5 h under white light and then incubated for 19 h in the dark at 25°C after exposure to the UVB doses of 0.02–0.5 J/cm². (D and E) Microscopic images (scale: 20 µm) and germination percentages of the tested strains’ conidia photoreactivated (5 h light + 19 h dark) or reactivated by 24-h dark incubation (NER) at 25°C after exposure to UVB irradiation of LD₉₅. *P < 0.001 in Tukey’s test. Error bars: SDs from three replicates.

Fig 4

Insight into divergent roles of Rad4A, Rad4B, and Rad23 in M. robertsii. (A and B) Y2H assays for Rad4A-Rad23 and Rad4A-Rad10 interactions. Note that the diploids AD-Rad4A-BD-Rad23 and AD-Rad10-BD-Rad4A (underlined) grew on quadruple-dropout plates (right in each two-image panel) like positive control (AD-LargeT-BD-P53) during 3-day incubation at 30°C. The colonies of each construct were initiated with 5 × 10⁴ (C1), 5 × 10³ (C2), and 5 × 10² (C3) cells, respectively. (C) Relative transcript (RT) levels of photorepair-depending (phr1, phr2, wc1, and wc2) or photoreactivation-required (rad1 and rad10) genes in the 3-day-old PDA cultures of rad4A mutants versus the WT strain. (D and E) RT levels of three central developmental pathway-activating genes and six catalase genes in the 3-day-old PDA cultures of rad23 mutants versus the WT strain, respectively. The dashed line indicates a significant level of onefold downregulation. (F) Total catalase activity assessed from the protein extracts of the tested strains’ cultures. (G) Concentrations of protoplasts released from cell suspensions of rad23 mutants and WT after 3, 6, and 9 h of cell wall lysing with enzymes in 1 M NaCl at 37°C. P < 0.05* or 0.001*** in Tukey’s test. Error bars: SDs from three cDNA samples analyzed per strain or three replicates.

Fig 5

Schematic diagram for links of Rad4A to enzymes/proteins required for photorepair and/or photoreactivation in M. robertsii. The multiple protein-protein interactions detected in the present (Rad4A-Rad10 and Rad4A-Rad23) and previous studies (37, 39) implicate that Rad4A is tied to a putative large protein complex that comprises two photolyases (Phr1 and Phr2), photolyase regulators (WC1 and WC2), and Rad1 and Rad10.