New insights into the evolution of subtilisin-like serine protease genes in Pezizomycotina

Abstract

Background: Subtilisin-like serine proteases play an important role in pathogenic fungi during the penetration and colonization of their hosts. In this study, we perform an evolutionary analysis of the subtilisin-like serine protease genes of subphylum Pezizomycotina to find if there are similar pathogenic mechanisms among the pathogenic fungi with different life styles, which utilize subtilisin-like serine proteases as virulence factors. Within Pezizomycotina, nematode-trapping fungi are unique because they capture soil nematodes using specialized trapping devices. Increasing evidence suggests subtilisin-like serine proteases from nematode-trapping fungi are involved in the penetration and digestion of nematode cuticles. Here we also conduct positive selection analysis on the subtilisin-like serine protease genes from nematode-trapping fungi.

Results: Phylogenetic analysis of 189 subtilisin-like serine protease genes from Pezizomycotina suggests five strongly-supported monophyletic clades. The subtilisin-like serine protease genes previously identified or presumed as endocellular proteases were clustered into one clade and diverged the earliest in the phylogeny. In addition, the cuticle-degrading protease genes from entomopathogenic and nematode-parasitic fungi were clustered together, indicating that they might have overlapping pathogenic mechanisms against insects and nematodes. Our experimental bioassays supported this conclusion. Interestingly, although they both function as cuticle-degrading proteases, the subtilisin-like serine protease genes from nematode-trapping fungi and nematode-parasitic fungi were not grouped together in the phylogenetic tree. Our evolutionary analysis revealed evidence for positive selection on the subtilisin-like serine protease genes of the nematode-trapping fungi.

Conclusions: Our study provides new insights into the evolution of subtilisin-like serine protease genes in Pezizomycotina. Pezizomycotina subtilisins most likely evolved from endocellular to extracellular proteases. The entomopathogenic and nematode-parasitic fungi likely share similar properties in parasitism. In addition, our data provided better understanding about the duplications and subsequent functional divergence of subtilisin-like serine protease genes in Pezizomycotina. The evidence of positive selection detected in the subtilisin-like serine protease genes of nematode-trapping fungi in the present study suggests that the subtilisin-like serine proteases may have played important roles during the evolution of pathogenicity of nematode-trapping fungi against nematodes.

Figure 1

Alignment of the cuticle-degrading protease sequences from nematophagous and entomopathogenic fungi. Areas shaded in black are conserved regions (100% similarity), areas shaded in grey have a high degree of homology (more than 75% similarity) and unshaded areas are regions of variability between the proteases. Signal peptide sequences were encompassed by frame with black edge. • indicates the aspartic acid (Asp₄₁)-histidine (His₇₇)-serine (Ser₂₃₁) (in EF113092) catalytic triad. C indicates the conserved cysteines. The underlined regions are the substrate-binding S₁ pocket in subtilisin-like serine protease gene. GenBank number: Arthrobotrys oligospora (PII), X94121; Arthrobotrys oligospora (Azol1), AF516146; Arthrobotrys conoides, AY859782; Arthrobotrys yunnanensis, EF113089; Arthrobotrys musiformis, EF113088; Arthrobotrys multisecundaria, EF055263; Monacrosporium cystosporum, AY859780; Monacrosporium elegans, AY859781; Monacrosporium coelobrochum, EF113091; Monacrosporium haptotylum EF681769; Monacrosporium megalosporum, AB120125; Monacrosporium microscaphoides, AY841167; Dactylella varietas, DQ531603; Dactylella shizishanna, EF113092; Monacrosporium psychrophilum, EF113090;Pochonia chlamydosporia, AJ427460; Lecanicillium psalliotae, AY692148; Paecilomyces lilacinus, EF094858; Hirsutella minnesotensis, EF560594; Hirsutella rhossiliensis, DQ422145; Beauveria brongniartii, AY520814; Beauveria bassiana, EF195164;Metarhizium anisopliae, M73795; Cordyceps sinensis (Csp1), EU282382; Cordyceps sinensis (Csp2), EU282383.

Figure 2

Phylogenetic tree based on amino acid sequences of subtilisin-like serine protease genes. The tree was constructed using MrBayes 3.1.2 [49]. The right frame diagrams show the intron distribution (position and phase) in corresponding proteins. SUB1-7 means seven genes encoding putative subtilisin-like serine proteases isolated from dermatophytic fungi.

Figure 3

Phylogeny based on subtilisin-like serine protease genes from nematode-trapping fungi used for ML analysis in PAML and Bn-Bs. Maximum parsimony (MP), neighbor-joining (NJ), maximum-likelihood (ML) and Bayesian tree reconstructions of the subtilisin-like serine protease gene sequences of clade D presented similar overall topologies. The bootstrap values of each branch for different methodologies are indicated (Bayesian/ML/NJ/MP). The symbol (*) indicates distinct topological arrangements. The thick branches indicate the branch with a significant LRT in the PAML branch analysis. After calculation using Bn-Bs, only branch f (Z = 3.66320) was significant.

Figure 4

The infection of the root-knot nematode Meloidogyne sp. with entomopathogenic fungi and nematode-parasitic fungi. (I), Arrow with "a" was the eggs of the root-knot nematode Meloidogyne sp. A, Beauveria bassiana; B,Metarhizium anisopliae; C,Lecanicillium psalliotae; D,Pochonia chlamydosporia; E,Paecilomyces lilacinus; F,Paecilomyces farinosus; G,Paecilomyces fumosoroseus; H,Metarhizium flavoviride; I,Nigrospora oryzae. A-F, H and I, bar = 30 μm. G, bar = 15 μm. All the eight tested entomopathogenic or nematode-parasitic fungi (A-H) were able to infect the eggs of the root-knot nematode Meloidogyne sp. (II), The histogram shows the infection rate of 3 days. After 1 week, the infection rate of eight tested entomopathogenic or nematode-parasitic fungi ranged from 73% to 100%. The phytopathogenic fungus N. oryzae (I-I) that used as the negative control did not show any infection toward the eggs, so the infection rate was zero.

Figure 5

The infection of the potato tuber moth P. opercullella with entomopathogenic fungi and nematode-parasitic fungi. (I), Arrow with "a" was the eggs of potato tuber moth P. opercullella. Arrow with "b" was the selected fungal strains. Strains A-I have been described in Figure 4. Each fungal culture is 6 mm in diameter. For the negative control (I-I) showed that all of the juveniles had hatched within 1 week. (II), Although the bioassay lasted for 1 week, the infection rate up to highest after 5 days later, so we only show the infection rate of 3-5 days. The infection rate of eight tested entomopathogenic or nematode-parasitic fungi ranged from 80% to 100%. The phytopathogenic fungus N. oryzae (II-I) that used as the negative control did not show any infection toward the eggs and the juveniles were all hatched into juveniles within 5 days, so the infection rate was zero.

Figure 6

Effects of the subtilisin-like serine protease PSP-3 on the nematode and insect eggs. (I), Effect of the subtilisin-like serine protease PSP-3 produced by P. lilacinus on the eggs of the root-knot nematode Meloidogyne sp. The average amount of protein released into the supernatant of the treatment with the active protease (a1), none (b1) and denatured protease (c1); (*, P < 0.05). (II), Effect of the subtilisin-like serine protease PSP-3 produced by P. lilacinus on the eggs of the potato tuber moth P. opercullella. The average amount of protein released into the supernatant of the treatment with the active protease (a2), none (b2) and denatured protease (c2); (*, P < 0.05).