C. higginsianum wild-type strain IMI349061 (CH-1) was cultured on potato dextrose agar (PDA) at 25°C and stored in PDA slants at 4°C for further use. The wild type plants of A. thaliana Col-0 used in this study for pathogenicity test were grown in greenhouse at 20 ± 2°C, under a 12 h light/dark cycle. Agrobacterium tumefaciems strain EHA105 was used for transformation of C. higginsianum and Escherichia coli strain JM109 was used to propagate all plasmids.

The publicly available genomic sequence database of C. higginsianum at JGI (http://genome.jgi.doe.gov/Colhi1/Colhi1.home.html) was used to characterize ChPKA1 (CH063_00098), ChPKA2 (CH063_12956), and ChAC (CH063_06008) genes. BLAST analysis was done by using NCBI (http://www.ncbi.nlm.nih.gov/) and UniProt (http://www.uniprot.org/blast/). MEGA program was used for the production of phylogenetic tree with unrooted neighbor-joining method. Domains identification was performed by using SMART MODE (http://smart.embl-heidelberg.de/smart/change_mode.pl). The Clustal X program was used for amino acid alignments.

The genomic DNA of C. higginsianum wild type strain and other derivative transformants were extracted using genomic DNA purification kit (Axygen, USA) according to the manufacturer's protocols. Southern blot analysis was performed as previously described (Liu et al., 2013). The 20 μg genomic DNA of each strain was digested overnight, then size-fractionated through a 0.8% agarose gel and mounted on positively charged nylon membrane. The nylon membrane was then hybridized with a probe amplified by primers (Supplementary Table S1) and labeled with digoxigenin (DIG)-dUTP using the PCR DIG Probe Synthesis Kit (Roche, Mannheim, Germany) according to the manufacturer's protocol.

Total RNA was isolated with TriZOL reagent (Invitrogen, Carlsbad, USA) according to manufacturer's instructions and then stored at −80°C for further studies. The total RNA samples were treated with DNase I (RNase Free) (Takara, Dalian, China) at 37°C for 30 min and used to generate the first strand cDNA with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems™, Foster, CA, USA) according to manufacturer's instructions. The first strand cDNA was stored at −20°C for further studies. Genes expression was analyzed by quantitative real-time reverse polymerase chain reaction (qRT-PCR) using a Bio-Rad CFX96 (Bio-Rad, California, USA) and SYBR Premix Ex Taq II (TAKARA, Dalian, China), according to the manufacturer's instructions. The C. higginsianum β-tubulin gene (CH063_04743) was used to normalize the RNA sample for each round of qRT-PCR. The PCR conditions were as follows: denaturation at 95°C for 3 min; 40 cycles of 95°C for 15 s, 55°C for 20 s and 72°C for 25 s; final step of 72°C for 10 min. Primer pairs for qRT-PCR reactions were designed either across or flanking an intron and listed in Supplementary Table S1. For each gene, qRT-PCR assays were repeated at least twice, with each repetition having three independent replicates.

To characterize ChPKA1, ChPKA2, and ChAC genes, the genes replacement vectors pChPKA1-3300, pChPKA2-3300, and pChAC-3300 were generated as described (Ma et al., 2017). The 5′(about 500 bp)- and 3′(about 500 bp)- flanks of the ORF of each genes were amplified from genomic DNA of the wild type strain CH-1 by PCR with primer pairs (Supplementary Table S1). The 5′- and 3′- fragments of each genes were then cloned into the upstream and downstream of hph cassette respectively, using Gibson Assembly Master Mix kit (New England Biolabs, Massachusetts, USA) according to the manufacturer's instructions. Then, the 5′ fragment-hph-3′ fragment cassettes of each gene were cloned into pNeoP3300 (Wei et al., 2016), resulting in gene replacement vectors, which had the selective marker hph gene flanked by the ORF flanking sequences from each of the genes (Supplementary Figure S1A).

The gene complementation vectors pChPKA1-Com, pChPKA2-Com and pChAC-Com were constructed that the cDNA of each gene was amplified by RT-PCR with primers (Supplementary Table S1) and cloned into the same sites of pCIT vector, which contained the constitutive PtrpC promoter and TtrpC terminator. Finally, the PtrpC-cDNA-TtrpC cassette was cloned into pNeoP3300, resulting in complementation vector.

Agrobacterium-mediated transformation was performed as previously described (Wei et al., 2016). Transformants were transferred to PDA plates containing 50 μg/ml of hygromycin B (Roche, Mannheim, Germany) or 150 μg/ml G418 (Ameresco, OH, USA) for a second round of selection and further confirmed by Southern blot and qRT-PCR analysis.

The growth rates of all the transformants and the wild-type strain were assayed as previously described (Wei et al., 2016). All strains were initially grown on PDA for 7 days. The mycelial agar discs were then taken from the active colony edge, inoculated into the center of the PDA petri dish at 25°C, and then the colony diameters from 1 to 7 dpi were examined. The colony morphology and conidiation of all strains were examined after being grown on PDA plates at 25°C for 14 and 7 days, respectively.

Pathogenicity assays of C. higginsianum on Arabidopsis were performed as previously described (Yuan et al., 2016). The conidia harvested from 5-days-old PDA agar cultures at 25°C were washed with sterile distilled water twice. The leaf surfaces of each pot of 4-weeks-old Arabidopsis were sprayed with 1 ml conidial suspensions (10⁶ conidia/ml). The infected Arabidopsis plants were incubated in humid chambers at 25°C for 6 days, and the symptoms were then observed. The experiment was repeated three times, and each strain infected three A. thaliana plants each time.

Appressorial formation was measured. 10 μl conidial suspensions (10⁶ conidia/ml for high conidial density and 10⁴ conidia/ml for low conidial density) were spotted on hydrophobic coverslips and incubated in darkness at 25°C for 24 h. The percentage of appressorial formation was determined by microscopic examination for at least 200 conidia or appressoria using Nikon Eclipse 80i microscope (Nikon, Tokyo, Japan), under bright-field model using 40 × fold magnification. Each test was repeated at least three times.

The tolerance of transformants and wild-type strain to multiple stress was performed. The mycelial plugs from the active colony edge were inoculated on PDA containing 300 μg/ml Calcofluor White (CFW), 10 mM H₂O₂, and 300 μg/ml Congo Red (CR) respectively and PDA media as controls. The growth inhibition rate (%) of stress to the transformants and wild type strain was analyzed by measuring the colony diameters as previously described (Wei et al., 2016).

The tolerance of these strains to high temperature was performed by inoculating mycelial plugs from the active colony edge of each strain on PDA and then culturing at 25 and 35°C respectively for 6 days. The phenotypes and growth inhibition rate of these strains were then analyzed.

The data were analyzed with Origin 7.5 (OriginLab Corporation, Massachusetts, USA) using ANOVA (one-way, P ≤ 0.01). Results of all graphs represent the mean value ± SD. Asterisks and different letters in the graphs indicate statistical differences, P ≤ 0.01.

The C. higginsianum genome contains two PKA catalytic subunit genes, CH063_00098 and CH063_12956, that were respectively named ChPKA1 and ChPKA2. ChPKA1 encodes a 507-amino-acids protein and ChPKA2 encodes a 392-amino-acids protein. The SMART MODE (http://smart.embl-heidelberg.de/smart/change_mode.pl) analysis result indicates that both of ChPKA1 and ChPKA2 contain a Serine/Threonine protein kinase catalytic (S_TKc) domain and an Extension to Ser/Thr-type protein kinase (S_TK_X) domain (Supplementary Figure S2A), which plays a key role in catalysis of protein phosphorylation. BLAST searches of fungal genomes with ChPKA1 and ChPKA2 showed the presence of homologs in a large number of fungi. Multiple sequence alignment and phylogenetic analysis revealed significant sequence conservation (Supplementary Figures S2B–D).

To gain insight into the functions of ChPKA1, ChPKA2, and ChAC, we firstly examined the gene expression patterns during different infection stages of C. higginsianum by qRT-PCR as previous described (Liu et al., 2013). The result demonstrated that the expression levels of ChPKA1 and ChAC but not ChPKA2, were significantly up-regulated during conidia germinating (5 hpi) and appressorial formation (22 hpi) stages (Figure 1), indicating that ChPKA1 and ChAC may play significant roles in conidia germinating and appressorial formation of C. higginsianum.

To assess the function of ChPKA1 and ChPKA2 in C. higginsianum, the ChPKA1 and ChPKA2 deletion mutants were generated and screened by growing on PDA containing hygromycin and further confirmed by Southern blot and qRT-PCR (Supplementary Figures S1B,C). Whereas the ChPKA2 deletion mutant ΔChPKA2 had no detectable phenotype variation compared with wild type strain, the ChPKA1 deletion mutant ΔChPKA1 showed a significantly reduced growth rate (Figures 2A, 3; Supplementary Table S2), and the ΔChPKA1 sectored on PDA with obvious dark color and abnormal colony (Figure 2A). Compared with ΔChPKA2 and wild type strain, the ΔChPKA1 also showed obvious reduction of conidiation (Figure 4; Supplementary Table S2). For complementation assays, the complementary vector pChPKA1-Com, pChPKA2-Com, and pChAC-Com were reintroduced into mutants to generate complementary strains ChPKA1-Com, ChPKA2-Com, and ChAC-Com. Defects in hyphal growth, colony morphology, and conidiation were rescued in ChPKA1-Com and ChAC-Com strains (Figures 2A, 3, 4; Supplementary Table S2). These results indicated that deletion of ChPKA1 but not ChPKA2 is directly responsible for the phenotype defects in ΔChPKA1, and ChPKA1 is involved in growth, colony phenotype and conidiation in C. higginsianum.

In addition, as adenylate cyclase (AC) contributes to the synthesis of cAMP and plays a significant role in cAMP-PKA signaling pathway, we also functionally analyzed ChAC by gene replacement of ChAC (CH063_06008) adenylate cyclase gene to generate ChAC knockout mutant ΔChAC (Supplementary Figure S1). Compared to wild type strain, the growth rate and conidiation of ΔChAC were obviously reduced on PDA medium (Figures 3, 4), and the ΔChAC also showed an albino colony after 6 days incubation on PDA (Figures 7A,B). Furthermore, in the presence of exogenous 10 mM cAMP in PDA medium, the growth rate and conidiation in the ΔChAC, but not ΔChPKA1, mutant were restored to the levels of the wild type strain (Figures 3, 4). These results further suggested the importance of cAMP-PKA signaling pathway in the regulation of growth and conidiation in C. higginsianum.

Totally, we had obtained at least five individual mutant strains and three individual complementary strains for each gene, and all the mutants and complementary strains of each gene showed the same phenotype (Supplementary Table S2). Thus, we selected two mutants and one complementary strain of each gene for Southern blot, and choose one mutant and one complementary strain of each gene for further phenotype analysis.

To further assess the function of ChPKA1 in pathogenicity, the infection assays against the Arabidopsis were performed as described above. The results demonstrated that the wild type, complementary strains and ΔChPKA2 could cause typically water-soaked and dark necrotic lesions on leaves, whereas the ΔChPKA1 and ΔChAC failed to cause obvious disease symptoms on Arabidopsis leaves after inoculating for 6 days (Figure 2B), which is similar to previous study in which the PKA and AC mutants of C. lagenarium did not cause any obvious lesions on cucumber leaves, even inoculated on wounded leaves (Yamauchi et al., 2004). These results indicated that ChPKA1 and ChAC are required for pathogenicity in C. higginsianum.

The appressorial formation of each strain was also analyzed. The results suggested that at high conidial density (10⁶ conidia/ml), most conidia (>65%) of wild type, complementary strains and ΔChPKA2 could form melanized appressoria on hydrophobic surface within 24 h (Figures 5A, 6A). In contrast, most ΔChPKA1 and ΔChAC conidia (>97%) did not form appressoria within 24 h. They showed obvious defects in appressorial formation and developed poorly (<3%) on hydrophobic coverslips (Figures 5A, 6A). The appressorium formation rates of ΔChPKA1 and ΔChAC were significantly lower than the rate of wild type strain CH-1 and complementary strains (Figure 6A). Interestingly, the appressorial formation defect in ΔChPKA1 and ΔChAC could be partially rescued by incubation of conidia at low density (10⁴ conidia/ml), but the appressorial formation rates were still much lower than the rate of wild type strain (Figure 6A). Although conidia of ΔChPKA1 and ΔChAC could form appressoria at low density (10⁴ conidia/ml), they developed much longer germ tubes compared with that of the wild type strain CH-1 (Figure 5B). In addition, application of exogenous cAMP could partly restore the size of the germ tube of ΔChAC mutant but not ΔChPKA1 (Figure 5B), and exogenous cAMP could also significantly increase the appressorial formation rate in ΔChAC but not ΔChPKA1 (Figures 6B,C). These results indicated that ChPKA1 and ChAC play significant roles in regulation of appressorial formation, thus to affect the pathogenicity of C. higginsianum.

To further examine whether the ChPKA1 and ChAC are involved in stress tolerance, the growth of wild type strain, ΔChPKA1, ΔChPKA2, and ΔChAC were measured on PDA plates supplemented with various compounds including 300 μg/ml Calcofluor White (CFW), 10 mM H₂O₂, and 300 μg/ml CR respectively. The tolerance of these strains to high temperature (35°C) was also determined. The results demonstrated that on PDA media with CFW or CR, inhibition of growth rate of ΔChPKA1 and ΔChAC was obviously higher than that of wild type strain and ΔChPKA2 mutant (Figures 7A,C), indicating that ChPKA1 and ChAC may function to maintain the fungal cell wall integrity by positive regulation.

In contrast, the ΔChPKA1 and ΔChAC mutants showed higher tolerance to 10 mM H₂O₂ and high temperature (35°C) compared with wild type strain (Figures 7B,D), which is similar with that of its homolog in F. verticillioides (Choi and Xu, 2010), suggesting that ChPKA1 and ChAC may negatively regulate H₂O₂ and high temperature tolerance in C. higginsianum. Moreover, previous studies reported that the GSY2 and HSP26 genes are up-regulated under heat stress condition in yeast (Schnell et al., 1992; Gasch et al., 2000), and the expression levels of FvGSY2 and FvHSP26 genes of F. verticillioides are higher in adenylate cyclase mutant after heat treatment (Choi and Xu, 2010). Similar results were also reported that the PKA mutant of F. graminearum was more tolerant to heat stress than the wild type strain for hyphal growth, and the relative expression levels of FgHSP70 and FgGSY2 in PKA mutant were much higher than in wild type strain when cultured at 35°C (Hu et al., 2014). In our study, we speculated that the homolog genes of GSY2 and HSP70 in C. higginsianum may be also involved in the high temperature tolerance of ΔChPKA1 and ΔChAC mutants. In order to verify this assumption, we identified the homolog of GSY2 in C. higginsianum, named ChGSY2 (CH063_00792), and homolog of HSP70, named ChHSP70 (CH063_01329). Then, we comparatively analyzed the expression levels of ChGSY2 and ChHSP70 in the hyphae of ΔChPKA1, ΔChAC, and wild type strain after incubating at 35°C for 3 h. The result demonstrated that the expression levels of ChGSY2 and ChHSP70 were obviously up-regulated in ΔChPKA1 and ΔChAC compared with that of the expression in wild type strain (Figure 8), indicating that ChPKA1 and ChAC may negatively regulate the expression of some stress response genes under high temperature condition in C. higginsianum.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.