Clg2p interacts with Clf and ClUrase to regulate appressorium formation, pathogenicity and conidial morphology in Curvularia lunata.

Ras is a small GTPase that regulates numerous processes in the cellular development and morphogenesis of many organisms. In this study, we identified and functionally characterized the Clg2p gene of Curvularia lunata, which is homologous with the Ras protein. The Clg2p deletion mutant (ΔClg2p) had altered appressorium formation and conidial morphology and produced fewer, smaller lesions compared with the wild-type strain. When a dominant Clg2p allele was introduced into the mutant, all of these defective phenotypes were completely restored. To further understand the regulation of Clg2p in appressorium formation and conidial morphology, and its role in pathogenicity, seven Clg2p-interacting proteins were screened using a yeast two-hybrid assay. Two of these proteins, Clf, a homologue of Mst11, which corresponds to MAP kinase kinase kinase in Magnaporthe oryzae, and urate oxidase (designated ClUrase) were functionally characterized. Clg2p specifically interacted with Clf through its RA domain to regulate appressorium formation and pathogenicity, whereas the Clg2p-ClUrase interaction regulated conidial morphology without affecting fungal pathogenicity. This report is the first to elucidate the regulatory mechanism of the key Ras protein Clg2p in C. lunata.

Curvularia leaf spot caused by Curvularia lunata (Wakker) Boed was responsible for one of most devastating diseases of corn in China during the 1990 s12. It is currently one of most widely distributed corn leaf diseases worldwide, resulting in substantial yield losses3456. The occurrence of various pathogenic types of C. lunata and its potential to evolve into other pathogenic strains has generated a global interest among scientists to uncover the molecular mechanism of its virulence and pathogenicity789101112. Understanding the molecular mechanism could reveal effective control strategies for this disease.

C. lunata mainly infects the leaves, sheaths and husks of corn plants. The initial yellow, water-soaked spots progress to round, spindle or oval shapes in later stages of infection. The distinctive features of this pathogen are dark brown conidiophores with characteristic boomerang or knee-like shapes on top. The conidia are crescent-shaped13. During the infection process, conidia of this pathogen land on the host plant and adhere to the leaf surface. Upon receiving appropriate stimuli during the initial host-surface recognition, signals are relayed for germ tube extension and production of infection structures, which then germinate to produce infection hyphae1415. The fungus also secretes lytic enzymes, which aid in penetrating the plant cell wall16. The germ tube may also differentiate into an appressorium after receiving appropriate physical or chemical signals from plant leaf surfaces, such as hydrophophobicity, topography or surface hardness, and then it uses mechanical force to penetrate the host plant. In C. lunata, the melanin present in the appressoria is also believed to reinforce mechanical strength during infection by maintaining turgor pressure17. The mechanism of penetration of C. lunata is similar to Magnaporthe oryzae18, but appressoria are not essential for penetration.

The interplay of signals between the host and pathogenic fungi is believed to be mediated by Ca2+, cyclic adenosine monophosphate (cAMP) and mitogen-activated protein kinase (MAPK)192021. In the past decade, MAPK has been increasingly implicated in regulating a host of functions in the fungus ranging from the development of hyphal growth, appressorium formation, maturation, invasive growth, conidiation and pathogenicity2223. For example, the Mst11-Mst7-Pmk1 MAPK cascade regulates appressorium formation and infectious growth of Magnaporthe grisea24. In Cochliobololus heterostrophus, Chk1 is involved in appressorium formation, female fertility and full virulence25. However, to date, only two homologues of MAPK, Clk and Clm1, have been identified in C. lunata. Clk is important for vegetative growth, biosynthesis of cell wall-degrading enzymes and pathogenicity10. The Clm1 mutant impairs fungus cell wall formation and conidial morphology, reduces conidiospore production, and lowers disease symptoms on corn leaves26. Although these findings expand our knowledge of the conidiation, cell wall formation and pathogenicity of C. lunata, the underlying molecular mechanisms regulating these processes are still not well understood. In our previous study, we created a normalized, full-length cDNA library of C. lunata and annotated a number of expressed sequence tag (EST) sequences using bioinformatics27. Among these ESTs, we identified a Ras homologue (designated Clg2p) from C. lunata and investigated its biological and regulatory role during infection.

The Ras protein family belongs to a class of small GTPases that are important organizers of signal transduction mechanisms due to their direct involvement in intracellular signal transduction pathways28293031323334. Ras proteins also affect cellular signal transduction pathways with important regulatory roles in morphogenesis, conidiation, appressorium development and pathogenicity in fungi35363738. For example, in Saccharomyces cerevisiae, Ras proteins elevate intracellular cAMP levels and trigger a kinase that results in sensitivity to heat shock, nutrient starvation, and inhibition of sporulation35. In Fusarium graminearum, the Ras2 protein reportedly affected the Gpmk1 MAP kinase pathway, regulating growth and pathogenicity of the fungus36. The Ras2 protein in Magnaporthe grisea is involved in appressorium formation and pathogenicity37. The Ras family homologous protein StRas2 in Setosphaeria turcica plays an important role in morphogenesis, conidiation, and appressorium development38. Little is known, however, about the role of Ras proteins in C. lunata.

The results presented in this study show that the Clg2p protein interacts with Clf and ClUrase, the urate oxidase protein, to regulate appressorium development and conidial morphology. This interaction is required for full pathogenicity in C. lunata. These results provide new insights into the signal transduction mechanism during the infection process of C. lunata and may lead to optimal target sites for chemical control of this devastating pathogen.

Results

Cloning and characterization of Clg2p

Among the various clones sequenced from the cDNA library, a particular sequence of 480 bp showed high similarity with Ras2p gene sequences of Pyrenophora tritici-repentis Pt-1C-BFP, indicating that it was most likely a partial sequence of Ras2p. This sequence has been designated Clg2p. Subsequent RACE reactions yielded a 797 bp fragment and a 175 bp fragment at the 3′-end and 5′-end, respectively. The complete full-length Clg2p contained a 678 bp open reading frame (ORF), a 61 bp 5′-untranslated region (5′-UTR) and a 549 bp 3′-UTR with a poly(A) tail. The ORF of Clg2p encoded a 225 amino acid protein with a molecular weight (MW) of 25.5 kDa and a calculated isoelectric point of 4.98. Using the ScanProsite tool, the encoded Clg2p protein was found to include 4 highly conserved GTP/GDP domains, a binding domain for the downstream effector molecule (RA), and a CAAX motif in the COOH-terminus, similar to other Ras proteins (Fig. 1A). A phylogenetic tree of Clg2p constructed using the neighbour-joining method indicated a close relationship with Ras2p from P. tritici-repentis Pt-1C-BFP (Fig. 1B). Analysis of the gene structure showed that Clg2p has four exons (53, 51, 386 and 176 bp) and three introns (58, 149 and 78 bp) within the 951 bp sequence (Fig. S1). Southern analysis identified a single copy of Clg2p in the C. lunata genome (Fig. 1C). The quantitative PCR (Q-PCR) analysis indicated that the expression level of Clg2p was significantly different at different developmental stages. The expression was highest in 3 h germinating conidia, indicating that the expression of Clg2p correlates with the pathogen infection, growth and appressorium development (Fig. 1D).

Figure 1

Molecular characteristics of Clg2p in C. lunata.

(A) Analysis of conserved domains of Clg2p. Box1, box3, box4 and box5 are the GTP/GDP domains; box2 is a binding site for downstream effector molecules (RA); and box6 is a variable region containing a CAAX motif. (B) A phylogenetic tree representing the phylogenetic relationships of Clg2p proteins and Ras proteins among related fungi. The number for each interior branch was the percentage of the bootstrap value (1000 replicates). (C) Analysis of copy number of Clg2p in the genome. A 951-bp PCR fragment amplified with primers 1F/R as a template of strain CX-3 DNA was labelled using Biotin to make the probe. (D) Expression patterns of Clg2p by qRT-PCR. M and C (x-axis) represent mycelial growth in potato dextrose (PD) medium for 3 d and conidia collected from 7 d culture on PDA plates at 28 °C in the dark, respectively. On the graph, 3 h, 6 h and 9 h (x-axis) represent germinating conidia collected from cellophane overlaid on PDA plates at 25 °C after 3, 6 and 9 h of growth, respectively. The error bars were calculated based on three replicates.

Targeted deletion and complementation of Clg2p

To generate the Clg2p deletion mutant, the Clg2p gene was replaced by the selective marker gene hph through ATMT transformation. The knockout construct was created by insertion 1090 bp upstream and 1094 bp downstream of Clg2p into the Δ1300 vector. This vector contains a selective marker gene, hph, flanked by the upstream and downstream sequences of the Clg2p ORF (Fig. 2A). The transformants were screened and selected on medium with 250 μg/ml hygromycin and verified by PCR for the presence of the resistance gene (hph). A genetic complementation strain (Clg2p-Com) was constructed by reintroducing the native Clg2p gene, including the 1.090-kb promoter region and DNA of Clg2p (Fig. 2B). The putative deletion mutants and genetic complementation strain were confirmed by Southern blotting (Fig. 2C,D) and RT-PCR (Fig. 2E). Finally, one deletion mutant and complementation strain were obtained and used for functional analysis in this study.

Figure 2

Construction and confirmation of the Clg2p deletion mutant and complementation strain.

(A) Construction of the deletion vector lacking Clg2p. (B) Construction of the complementation vector of Clg2p as described in the methods section. (C,D) Southern blotting analysis of the Clg2p deletion mutant (ΔClg2p) and complementation strain (Clg2p-Com). (C) Southern blotting was performed on isolated total genomic DNA from CX-3, ΔClg2p and Clg2p-Com after digestion with XbaI and BamHI. A 607-bp PCR fragment of hph was used as the hybridization probe. (D) A second Southern blotting analysis was performed using isolated total genomic DNA from wild-type strain CX-3, ΔClg2p and Clg2p-Com after digestion with a single-cutter restriction enzyme, HindIII, using a 951-bp PCR fragment of Clg2p as the probe. (E) Clg2p expression in CX-3, ΔClg2p and Clg2p-Com was analysed by RT-PCR using the Clg2p gene-specific primers clg2pds and clg2pda. A predicted 415-bp fragment was obtained from the wild-type CX-3 strain and the complemented strain but was absent in the ΔClg2p mutant.

The Clg2p gene affects appressorium formation and conidial morphology and is essential for its full pathogenicity

The results on colony morphology, mycelial vegetative growth, conidiation, conidia germination, conidial morphology and appressorium formation in the wild-type strain CX-3, the deletion mutant (ΔClg2p), and the complementation strain (Clg2p-Com) as well as their effect on pathogenicity are presented in Fig. 3A and Table 1.

Figure 3

Functional analysis of the Clg2p gene.

(A) The colony morphology of the wild-type strain CX-3 (WT), ΔClg2p and Clg2p-Com cultured on PDA media for 7 d; (B) Abnormal conidial morphology was observed by microscopy, bar = 20 μm; (C) Abnormal appressoria were observed by microscopy, bar = 20 μm; (D,E) Disease symptoms on unwounded (D) leaves of maize inoculated with mycelial plugs from CX-3, ΔClg2p and Clg2p-Com. Inoculation using agar plugs without fungus was used as a control. (E) Disease symptoms on susceptible corn leaves sprayed with conidial suspensions of CX-3, ΔClg2p and Clg2p-Com. Inoculation with sterile distilled water containing 2% glucose and 0.02% Tween 20 was used as a negative control. Typical leaves were photographed 3 d after inoculation.

Table 1

Phenotype of the mutant compared with the wild-type isolate CX-3 (WT).

Strain	Growth rate (mm/day)I	Conidiation (106 conidia/ml)II	Percentage of crescent- shaped conidia (%)III	Lesion areas (5 cm leaf tips)IV
CX-3 (WT)	7.45 ± 0.05a	7.37 ± 0.35a	90.56 ± 3.0a	174.41 ± 5.57a
ΔClg2p	7.05 ± 0.12d	5.41 ± 0.32c	36.1 ± 2.02d	42.49 ± 4.23d
Clg2p-Com	7.22 ± 0.09c	5.81 ± 0.42b	81.56 ± 3.47b	173.89 ± 6.89a
ΔClf	7.01 ± 0.08d	0 ± 0d	–	105.71 ± 7.37c
ClfΔRA	7.31 ± 0.08b	5.5 ± 0.18c	81.33 ± 3.80b	113.57 ± 5.59b
ΔClUrase	7.39 ± 0.05ab	5.94 ± 0.43b	43.33 ± 3.00c	171.90 ± 4.72a
Ect	7.32 ± 0.12b	5.86 ± 0.31b	82.67 ± 2.60b	172.21 ± 5.28a

IGrowth rate was measured on PDA medium.

IIConidial numbers harvested from a PDA plate (diameter = 9 cm) after a 12-d incubation at 28 °C.

IIIConidial morphology is expressed as the percentage of conidia that are crescent-shaped.

IVLesion areas of detached wounded leaves.

All data in all columns are the means of three independent experiments with standard deviations. The statistical analysis was performed using the SAS statistical package (Cary, North Carolina, USA). Statistically significant analysis of variance was further analysed using least significant difference tests. Different letters in each data column indicate significant differences at P = 0.05.

The most striking finding was the deformed and abnormal appressorium structure observed in ΔClg2p (Fig. 3B). By contrast, the wild-type strain CX-3 and Clg2p-Com showed normal appressorium structure and position. When we examined conidiation and conidial morphology, we found that most of the conidia produced by ΔClg2p were abnormal in appearance and mostly appeared as straightened conidia (Fig. 3C). However, the amount of conidia produced did not significantly change between ΔClg2p and the ectopic transformant (Ect). Clg2p-Com also showed normal conidial morphology. The pathogenicity of the deletion mutant was analysed by infection assays in the susceptible corn inbred line Huangzao 4. The results indicated that the lesions on unwounded leaves caused by ΔClg2p were smaller in shape than those caused by the wild-type strain CX-3 and Clg2p-Com (Fig. 3D). Moreover, in spray infection assays with 15 d old seedlings, numerous curvularia spots were observed on corn leaves sprayed with CX-3 or Clg2p-Com at 3 dpi, whereas fewer spots were observed on leaves sprayed with the ΔClg2p under the same experimental conditions (Fig. 3E). These results clearly suggest that Clg2p is required for full pathogenicity in C. lunata. In addition, our results also showed that the Clg2p gene has a role in appressorium formation and position and in conidial morphology.

Screening and identification of proteins associated with Clg2p

To understand the mechanism of Clg2p-mediated appressorium formation, conidial morphology and pathogenicity in C. lunata, we screened the constructed cDNA library of C. lunata for Clg2p-interacting proteins in a yeast two-hybrid (Y2H) assay. Before performing the Y2H assay, it was confirmed that pGBKT7-Clg2p was not activated automatically (Fig. S2) and does not have any toxic effects. The Y2H screen was performed twice on the cDNA library, and 7 proteins were obtained in total (Table S1). One of the Clg2p-interacting proteins is a homologue of MAPKKK, similar to Raf in S. Cerevisiae and Mst11 of the Mst11-Mst7-Pmk1 MAPK cascades in M. oryzae; this protein was designated Clf in this study. Another interacting protein was identified is urase oxidase and named ClUrase. Other proteins identified included membrane-binding proteins and hypothetical proteins, which are listed in Table S1. The Clf and ClUrase genes were selected for further characterization. The ORFs of the Clf and ClUrase sequences were extracted from C. lunata genomic DNA and confirmed by RT-PCR. The structural characteristics of the Clf and ClUrase proteins are shown in Figs S3 and S4A,B.

Interaction of Clg2p and Clf

To validate the interaction between Clg2p and Clf, we performed a targeted yeast two-hybrid (Y2H) assay, protein pull-down in vitro and a bimolecular fluorescence complementation (BiFC) assay. The results from the targeted yeast two-hybrid assay clearly showed that Clg2p physically interacted with Clf. In addition, the presence of a well-conserved Ras-association (RA) domain further supports the occurrence of such an interaction. Previous studies also indicated that such domains could interact with Ras-homologous proteins37, suggesting that the Clg2p protein may also interact with the RA domain of the Clf protein. Therefore, to further investigate whether the Clg2p protein interacts with the RA domain of Clf, we used a targeted yeast two-hybrid assay. The results shown in Fig. 4A clearly suggest that the Clg2p protein interacts with the RA domain of Clf but not the SAM domain of Clf.

Figure 4

The interaction between Clg2p and Clf.

(A) The interaction between Clg2p and Clf structural domains (RA and SAM) of Clf in yeast. Targeted H2Y with Clg2p as the bait and with the ORF, RA domain, or SAM domain of Clf as the prey. The interaction of pGBKT7-53 and pGADT7-T was used as the positive control, and the interaction of pGBKT7-Lam and pGADT7-T was used as the negative control. (B) A pull-down assay analysing the interaction between Clg2p and Clf. GST-Clg2p bound to GST beads were mixed with His-Clf purified from IPTG-induced E. coli BL21 (DE3). The bound proteins were eluted after removing the unbound proteins by washing and analysed by immunoblotting with anti-GST or anti-His antibodies. (C) Interaction between Clg2p and Clf in a plant cell. BiFC of Clg2p and Clf in transiently transformed onion epidermal cells. pspYNE-Clg2p+pspYCE, pspYNE+pspYCE-Clf, pspYNE-Clf+pspYCE and pspYNE+pspYCE-Clg2p were used negative controls. The YFP fluorescence signal was detected and observed using confocal microscopy.

For pull-down assays, the full length of Clg2p was expressed as a glutathione S-transferase (GST) fusion protein (GST-Clg2p) and was bound to a glutathione-agarose column. Additionally, Clf was also fused with the 6× His tag (His-Clf) and added to a glutathione-agarose column inoculated with GST-Clg2p, which was then eluted repeatedly with buffer to remove the non-specifically bound proteins. Proteins retained on the beads were eluted with buffer and resolved on SDS-PAGE (Fig. S5). The GST-Clg2p and His-Clf retained on the beads were detected by immunoblotting with anti-GST-tag and anti-His-tag antibodies, respectively. The results clearly suggest a direct interaction between the Clg2p and Clf proteins as shown in Fig. 4B. Furthermore, the BiFC assay allowed for visualization of the interactions to be mainly localized in the plasma membrane of the plant cells. As expected no YFP signals were observed in whole plant cells used as negative controls (Fig. 4C).

Clg2p physically interacts with Clf via its RA domain to regulate appressorium formation, and Clg2p affects pathogenicity

The above results indicated that Clg2p interacted with Clf, but only with the RA domain of Clf. To further investigate the functional relationship of Clg2p with Clf, we generated mutants with Clf ORF (ΔClf) and Clf-RA domain deletions (ClfΔRA) (Fig. 5A). The (ΔClf) and Clf-RA domain deletion (ClfΔRA) mutants were confirmed by Southern blotting (Fig. S6) and RT-PCR (Fig. S7). Interestingly, the ΔClf deletion mutant of C. lunata did not produce conidia on PDA medium (Fig. 5B). Most exhibited abnormal morphology such as the lack of conidia production or loss of function such as lost capacity to infect maize leaves, similar to the phenotype of the Clk deletion mutant observed in C. lunata19. A swollen structure formed in the hyphal tips of ΔClf (Fig. 5C), whereas, in the ClfΔRA mutant, most appressoria showed abnormal morphology and position compared with the wild-type strain CX-3 (Fig. 5C). Furthermore, smaller and fewer lesion areas were observed when either mycelial plugs or conidial suspensions of ClfΔRA mutants were sprayed to inoculate maize leaves than those observed when inoculated by the wild-type strain CX-3 (Fig. 5D,E). Overall, our results suggest that, in C. lunata, the interaction of Clg2p with Clf occurs via its RA domain, and this interaction is involved in appressorium formation, morphological development and pathogenicity.

Figure 5

Conidiation, appressorium formation and pathogenicity of the Clf deletion and RA domain deletion mutants.

(A) Construction of the Clf deletion and RA domain deletion mutants. (B) Abnormal conidial morphology was observed by microscopy, bar = 20 μm. (C) Abnormal appressorium structure and position were observed by microscopy, bar = 20 μm. (D) Leaves of maize inoculated by mycelial plugs from the wild-type strain CX-3, ΔClf and ClfΔRA. Inoculation using agar plugs without fungus was used as a control. (E) Disease symptoms on leaves of maize inoculated by conidial suspensions from the wild-type strain CX-3 and ClfΔRA. Inoculation with sterile water containing 2% glucose and 0.02% Tween 20 was used as a negative control.

Clg2p interacts with ClUrase to regulate conidial morphology

To confirm whether Clg2p also directly interacts with ClUrase and, if so, to determine the location of their interactions, we performed the same set of assays as those described above. The results showed that a positive interaction between Clg2p and ClUrase in targeted Y2H assays (Fig. 6A). In the pull down assays, the His-ClUrase fusion protein was successfully induced to express in E. Coli BE3, and inoculated into the beads containing GST-Clg2p. After elution of non-specificially protein, the bait and prey protein were also eluted (Fig. S8), and subjected to Western blotting analysis. The results showed that the GST-Clg2p and His-Clf were detected by immuno-bloting with anti GST-tag and His-tag antibody respectively (Fig. 6B), confirming the interaction in-vitro between Clg2p and ClUrase. The BiFC assay further indicated that the interaction was mainly restricted to the cell membrane and the nucleus of the plant cell (Fig. 6C).

Figure 6

Clg2p interacts with ClUrase.

(A) Clg2p interacts with ClUrase in yeast cells. The full-length gene ClUrase was used as bait and the full-length gene Clg2p was used as prey in targeted Y2H assays. (B) Clg2p directly interacts with ClUrase in vitro. In the pull-down assays, GST-Clg2p bound to GST beads was mixed with His-ClUrase purified from IPTG-induced E. coli BL21 (DE3). The bound proteins were eluted after removing unbound proteins by washing and analysed by immunoblotting with anti-GST or anti-His antibodies. (C) BiFC analysis of the Clg2p interaction with ClUrase. Clg2p and ClUrase were constructed with the N-terminal (YN) and C-terminal (YC) halves of YFP separately under the control of a CaMV35S promoter and were co-expressed transiently in onion epidermal cells. YFP expression was monitored by confocal laser scanning microscopy. pspYNE-Clg2p+pspYCE, pspYNE+pspYCE-ClUrase, pspYNE-ClUrase+pspYCE and pspYNE+pspYCE-Clg2p were used as negative controls. The YFP signal was observed using confocal microscopy.

To determine the functional relationship of the Clg2p and ClUrase interaction, we generated a ClUrase deletion mutant (designed ΔClUrase) by transforming it in hygromycin-resistant protoplasts of C. lunata (Fig. 7A). PCR analysis from three of the transformants suggested that only one may be a ClUrase gene deletion mutant because it had the hph gene but lacked the ClUrase gene. Southern analysis (Fig. 7B,C) and RT-PCR (Fig. 7D) further confirmed that this transformant lacked the ClUrase gene compared with the wild type. Although ΔClUrase had no obvious changes in colony morphology, vegetative growth, conidial yield or appressorium formation (Fig. 7E), we observed a clear change in the conidial morphology of the ΔClUrase mutant, with approximately 47.23% uncurved conidia, compared with the wild-type strain (Fig. 7F and Table 1). To determine whether these abnormal conidia affected the pathogenicity of the organism, conidial suspensions of the wild-type CX-3 and ΔClUrase strains were inoculated on unwounded maize leaves. Inoculation with ClUrase and the wild-type strain CX-3 resulted in the same size lesions, suggesting that the ClUrase disruption had little or no effect on pathogenicity (Fig. 7G). Therefore, our results suggest that the interaction of Clg2p with ClUrase apparently regulated the conidial morphology without affecting the pathogenicity of the pathogen.

Figure 7

Construction and confirmation of the ClUrase deletion mutant and functional analysis of ClUrase.

(A) Construction of the deletion vector for ClUrase. Southern blotting analysis of the ClUrase deletion mutant (ΔClUrase). Total genomic DNA from CX-3 and ΔClUrase were digested with the restriction enzyme PaeI and probed with a 610-bp fragment of hph (B). Southern blotting with a 688-bp PCR fragment of ClUrase used as the probe (C). RT-PCR analysis of ClUrase expression using the ClUrase gene-specific primers ClU-F and ClU-R in the wild-type strain CX-3 and ΔClUrase. A predicted 654-bp fragment was found in the wild-type strain CX-3 but was absent in the ΔClUrase mutant (D). Microscopic observation of appressorium formation and position, bar = 20 μm (E). Conidial morphology was observed by microscopy, bar = 20 μm (F). Leaves of maize inoculated with mycelial plugs from CX-3 and ΔClUrase. Inoculation using agar plugs without fungus was used as a control (G).

Discussion

C. lunata is a phytopathogenic fungus that causes Curvularia leaf spot in maize. After a serious outbreak in the 1990 s, scientist achieved moderate success in controlling the fungus with the introduction of resistant maize cultivars. However, the disease has returned in recent years due to the continuous cultivation of these cultivars. This selection pressure is believed to have led to the evolution of the pathogen, which produces a highly virulent strain of C. lunata. Therefore, the pathogen poses a major threat to the agro-security of China122139. A priority in the current corn breeding programme is to understand the variability in virulence and pathogenicity factors in the plant-pathogen interaction, which are closely related, and then to develop resistant cultivars4041. To further this understanding, we developed a full-length cDNA library from the mycelium of C. lunata27. Bioinformatic analyses of the derived EST sequences identified a Ras homologous gene, Clg2p, with a close relationship to Ras2p from P. tritici-repentis Pt-1C-BFP. Initially, the expression pattern of Clg2p was highest during conidial germination, correlating its expression with pathogen infection and underscoring its potential role in fungal pathogenicity. To clarify the biological role of this gene during pathogenesis, we isolated the complete ORF and created its deletion mutant. The deletion mutant ΔClg2p was compared with the wild-type strain and its complemented strain for conidial morphology, mycelial growth, conidiation, appressorium formation and pathogenicity on susceptible corn plants. The most striking effects of the Clg2p deletion were deformed, abnormal appressoria and conidia. When tested for pathogenicity, the mutant produced fewer, smaller lesions than those of the wild-type strain or its complemented strain. These effects were reversed when the native dominant gene was replaced in the mutant. This result suggested that Clg2p is an important component of a signal transduction pathway involved in the pathogenicity of C. lunata. Therefore, to identify the complete network of other interacting proteins of Clg2p, we used yeast two-hybrid screening. Two proteins in particular, Clf and ClUrase, were selected for further characterization based on the bioinformatics analysis. Clf was identified as a homologue of Mst11 from M. oryzae. A targeted yeast two-hybrid assay, protein pull-down in vitro and a bimolecular fluorescence complementation assay not only revealed that Clg2p interacted physically with Clf via its RA domain but also showed that the interaction was localized to the plasma membrane of the plant cell. Interestingly, the mutant generated for the domain (ClfΔRA) did produce conidia but the appressoria developed from them were abnormal in appearance. In the pathogenicity assays, the mutants produced fewer, smaller lesions on inoculated leaves. These results were similar to those of ΔClg2p described above. A functional analysis of ClUrase also revealed a physical interaction with Clg2p, which was restricted to the cell membrane and the nucleus. Moreover, the ClUrase mutant (ΔClUrase) produced abnormally shaped conidia similar to ΔClg2p. Unlike ΔClg2p, however, the pathogenicity of ΔClUrase was not affected. Overall, our results suggested that Clg2p interacted with Clf to regulate appressorium formation, morphology, and pathogenicity. However, Clg2p interacted with ClUrase to regulate conidial morphology without affecting the pathogenicity of C. lunata (Fig. 8).

Figure 8

Hypothetical model of functions of Clg2p in Curvularia lunata.

Clg2p is a Ras GTPase protein that interacts with the RA of MAPKKK (Clf). Clg2p acts with Clf to regulate the MAPK pathway and affects appressorium formation and pathogenicity in Curvularia lunata. Additionally, Clg2p interacts with ClUrase, which affects conidial morphology.

The Ras protein is a key regulatory molecule and is well conserved across kingdoms. It acts as a switch by cycling between inactive and active GTP-bound forms. It plays an important role in controlling cell growth, proliferation, and morphogenesis4243. Recently, there has been a growing interest in studying the role of Ras proteins in fungal development, morphogenesis, and pathogenesis44454647484950. In some well-studied plant fungal pathogens, such as M. oryzae37, S. turcica38 and Aspergillus nidulans46, knocking out the Ras2 gene either affected the formation of appressoria or caused a variety of morphological defects, such as slower growth and delayed spore germination. In other studies, mutations within Ras either eliminated pathogenicity or caused significant reductions in virulence364850. These observations are consistent with our findings that Clg2p was essential for the development and pathogenicity of C. lunata.

Based on our findings, we presented a model of the involvement of Clg2p in the Ras signalling pathway-mediated appressorium and conidial development and its effect on the pathogenicity of C. lunata (Fig. 8). Our results showed that Clg2p physically interacted with Clf through its RA domain, which acts upstream of Clf to regulate appressorium formation and pathogenicity. Conversely, the interaction of Clg2p-ClUrase altered conidial morphology severely without affecting pathogenicity (Fig. 8). Clg2p is a homologue of the Ras protein family and is known to interact with downstream proteins in the MAPK and cAMP pathways that regulate cellular development and morphogenesis. In S. cerevisiae35, for example, Ras2p regulates invasive fungal growth by interacting with Cdc42 from the MAPK pathway and Cyr1 from the cAMP pathway. In F. graminearum, Ras2 regulates pathogen growth and virulence through the Gpmk1 MAP kinase pathway36. The Ras2 of U. maydis, however, promotes bud growth by activating the cAMP pathway and regulates morphogenesis by signalling through a MAP kinase cascade48. In our study, Clg2p was also an important component of the MAPK cascade pathway, and it regulated pathogenicity and the formation of appressoria in C. lunata.

We annotated all MAPK pathway homologues in C. lunata using information from its genome (unpublished) and identified two key members of the Clf-mediated MAP kinase cascade, MAP kinase kinase and MAP kinase (Clk1). This pathway is similar to the Fus/Kss1-MAPK pathway of S. Cerevisiae and the Mst11-Mst7-Pmk1 cascade of M. grisea37. Interestingly, knocking out Clf (Mst11 homologue) and Clk1 (Pmk1 homologue) in C. lunata led to a failure to produce conidia, suggesting that they play roles in regulating asexual reproduction and in the production of conidia. By contrast, Mst11 and Pmk1 deletion mutants could produce conidia and failed to form appressoria37. This novel finding suggests that the MAPK pathways in these two fungal pathogens have slightly different roles. These results will direct future experimentation designed to reveal the regulatory mechanism of conidiation by the Clf-mediated MAPK pathway in C. lunata.

Our identification of the protein urase oxidase and how its interaction with the Clg2p protein affected conidial morphology has not been reported in previous studies. However, urase oxidase was identified in the interaction between Fusarium oxysporum and tomato. Urase oxidase can catalyse the conversion of uric acid to allantoin, forming NH4+ ions that may be a nitrogen source for the fungus51. In microorganisms that infect humans, the ability to hydrolyse urea is linked to several consequences, such as formation of urinary stones in humans pyelonephritis, peptic ulceration and hepatic encephalopathy52. Interestingly, C. lunata is reported to cause human respiratory tract, skin and corneal infections5354. Therefore, we hypothesized that deletion of the C. lunata urase oxidase gene affected the utility and absorption of uric acid and altered conidial morphology.

In conclusion, we identified and characterized the Clg2p gene in C. lunata and found it to be similar to the Ras gene in other phytopathogenic fungi. We proposed its participation in appressorium formation, conidial morphology, and pathogenicity. In addition, we demonstrated participation of the Clf-MAPK cascade in conidiogenesis by the lack of conidia formation in the Clf and Clk1 mutants. Further investigation is required to define the regulatory mechanism of conidiation in this pathogen. The results of this work may provide target sites for designing new chemicals to manage C. lunata and similar fungi.

Materials and Methods

Fungal strains and culture conditions

C. lunata CX-3 provided by Prof. Chen (Shanghai Jiaotong University, Shanghai, China) was used as the wild-type strain and was maintained on potato dextrose agar (PDA) at 28 °C. The selective medium was supplemented with either 250 μg/ml of hygromycin B (50 mg/ml, Roche Applied Science) or 200 μg/ml of Neomycin (Sigma-Aldrich Co. LLC) according to the selection marker gene.

Cloning full-length cDNA and DNA of Clg2p

In our blast analysis, the EST identified as Clg2p showed high similarity with Ras2p from Pyrenophora tritici-repentis Pt-1C-BFP, which was 480 bp in length. To obtain the full-length cDNA sequence of the Clg2p gene, 5′-RACE and 3′-RACE were performed using SMARTTM RACE cDNA Amplification Kits (Clontech, CA, USA) according to the manufacturer’s instructions. For 5′ RACE PCR amplification, the first and second round PCR reactions were amplified with ClR-5 and UPM, NUP respectively. The cDNA 3′-end was obtained using ClR-3 and a 3′ site adaptor primer. To determine the nucleotide sequence of genomic DNA corresponding to Clg2p cDNA, one specific primer pair (1F/R) was designed according to the identified ORF of Clg2p. The PCR product was sequenced after cloning into the pMD19-T vector.

Sequence, phylogeny, copy number and expression analysis of Clg2p

Nucleotide and deduced amino acid sequences were analysed using ClustalX1.83 software. Molecular weight determination and isoelectric point predication were conducted by ExPASy. Compute pI/Mw tool (http://www.expasy.ch/tools/pi_tool.html), and conserved domais of the deduced protein were determined using the ScanProsite tool. The phylogenetic tree was constructed using the neighbour-joining method. To assay gene copy number in the genome and confirm the deletion mutants, Southern blotting was performed as previously described by Liu et al.55.

Real-time quantitative PCR was used to analyse Clg2p gene expression at different growth stages as previously described by Liu et al.56. Primers 2F/R were used for Clg2p gene amplification. GAPDH was used as an endogenous control. The relative quantitative transcripts were calculated using the 2−ΔΔCt method.

Clg2p gene replacement, mutant and complementation

To replace the Clg2p gene, first, cassette polymerase chain reaction was used to rescue two flanking sequences of the Clg2p gene and was performed using the LA PCR in vitro Cloning Kit (Takara Bio Inc., Japan) according to the manufacturer’s instructions. Briefly, a DNA sample from CX-3 was digested with EcoRI, HindIII, PstetcI, SalI and XbaI restriction enzymes. The digested DNA was ligated to corresponding sites in the cassette (Takara Bio Inc., Japan). The cassette primers provided in the cloning kit and specific nested primers (3F/3R and 4F/4R) were used to amplify the upstream and downstream sequences of Clg2p. The deletion vector was constructed based on the two flanking sequences of Clg2p. A 1090-bp upstream fragment of Clg2p in the C. lunata genome was amplified with primers 5F/R and cloned between the HindIII and XbaI sites on Δ1300 plasmids, resulting in the p1300 g1 vector. Then, a 1094-bp downstream fragment of Clg2p was amplified with primers 6F/R and cloned between the BamHI and EcoRI sites on the p1300 g1 vector to create a Clg2p gene deletion vector (p1300 g11). The ATMT method was used to create the transformants and was performed as previously described by Liu et al.57.

A 2.112-kb fragment containing the 1.090-kb native promoter and ORF of Clg2p was amplified using primers 7F/R and was inserted between the HindIII and BamHI sites of pCAMBIA1300th, resulting in the complementation vector pCAMBIA1300th-Clg2p. The vector pCAMBIA1300th-Clg2p was co-transformed with the Clg2p deletion mutant followed by screening for Neomycin resistance.

Yeast two-hybrid screening and sequence analysis

Yeast two-hybrid (Y2H) screening was performed using the Matchmaker™ GAL4 Two-Hybrid System 3 & Libraries User Manual (Clontech). First, the open reading frame (ORF) of Clg2p was amplified with primers 8F/R and inserted into the EcoRI and BamHI sites of the pGBKT7 yeast vector (Clontech) to create a bait. An AD fusion library in pGADT7-rescuing cDNA from C. lunata was constructed and transferred into the yeast strain AH109 as described in Jing et al.58. The SD/-Ade/-His/-Leu/-Trp/X-α-Gal medium was used to screen and confirm expression. Then, plasmid DNA from the positive clones was isolated from yeast, transformed into E. coli and confirmed by sequencing.

Sequences from positive clones were analysed using BLASTx (http://www.ncbi.nlm.nih.gov). The predicted ORFs of positive colonies were extracted from the genome of C. lunata (http://www.ncbi.nlm.nih.gov/nuccore/JFHG00000000) and confirmed by RT-PCR with cDNA as a template. The percentage of similarity and identity of the known sequences were determined using the MatGAT programme. The conserved domains of the deduced proteins were analysed by CDD Search (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and Pfam (http://pfam.xfam.org/).

Target yeast hybrid assays

To confirm one-to-one interaction, the pGBKT7-Clg2p plasmid was used as bait. The ORF of Clf amplified using primers 9F/R, the SAM domain of Clf amplified using primers 9F and 10R, the RA domain of Clf amplified using primers 11F/R and the ORF of ClUrase amplified using primer 12F/R were inserted into the NdeI and BamHI sites of the yeast vector pGADT7 (Clontech) to create prey.quadruple dropout medium with -Trp/-Leu/-Ade/-His/X-α-Gal/AbA (QDO/X/A) that was used to detect the interaction. Co-transformants containing (pGBKT7-53+pGADT7-T) and (pGBKT7-Lam+pGADT7) vectors were used as positive and negative controls, respectively.

Pull-down and Western blotting

The ORF of Clg2p was amplified with primers 13F/R and inserted into pGEX-6P-1 to create a GST-Clg2p bait protein. The ORFs of Clf and ClUrase were amplified using primers 14F/R and 15F/R, respectively, and inserted into pET-28a vectors to create a His-Clf/ClUrase prey protein. The BL21 (DE3) strain was used to express the protein. In vitro pull-down assays were performed following the manufacturer’s instructions (Pierce™ GST Protein Interaction Pull-Down Kit (#21516, http://www.thermoscientific.com)). The bait and prey were eluted from 25 μL glutathione agarose. Pull-down proteins were then resolved by 12% SDS-PAGE and detected by Western blotting. GST-Clg2p and HIS-Clf/His-ClUrase fusion proteins were detected by immunoblotting with 1:2,000 diluted anti-GST antibody (#10000-0-AP, PROTEINTECH, http://www.ptgcn.com) and 1:1,500 diluted anti-His (#10001-0-AP, PROTEINTECH) antibody (Millipore), respectively. The goat anti-rabbit IgG (peroxidase conjugate) secondary antibody (Boster, http://www.boster.com.cn) was detected using DBA Western Blotting Substrate (Boster, http://www.boster.com.cn) following the manufacturer’s protocol.

Bimolecular fluorescence complementation assays

For BiFC assays, ORF of Clg2p,Clf and ClUrase were amplified with primers 16F/R, 17F/R and 18F/R, subcloned into pSPYNE-35S/ pSPYCE-35S (for split YFP N/C-terminal fragment expression) to form the fusion vector Clg2p-YNE/Clg2p-YCE,Clf-YCE/Clf-YNE and ClUrase-YCE/ClUrase-YNE. The pair of fusion proteins were transiently co-expressed into onion epidermal cells by the GV3101 strain (Agrobacterium tumefaciens) transfection as was described for subcellular localization of Clg2p genes, with transformed Clg2p-YNE, Clg2p-YCE, Clf-YNE, Clf-YCE, ClUrase-YNE and ClUrase-YCE. A confocal laser scanning microscope (Zeiss LSM510, Germany) was used to detect the fluorescence signal. Co-expressions of target genes and pspYNE or pspYCE were used as negative controls experiments.

Clf, ClUrase and RA domain of Clf deletion (Clf  ΔRA)

The double joint PCR technique was employed to construct genes deletion cassettes for Clf and ClUrase. The upstream and downstream sequences of both of genes were amplified based on the genome sequence of C. lunata. For Clf gene deletion, a deletion cassette containing the two flanking sequences of Clf surrounding hygromycin B phosphotransferase (hph) gene as a selectable marker was amplified by third round PCR described as Yu et al.59. In first-round PCR, three fragments including upstream fragment of Clf, downstream fragment of Clf and selectable marker (hph) with TrpC promoter of Aspergillus nidulans were amplified using primers C1F/C1R, C2F/C2R and Hph-F/Hph-R respectively. In second-round PCR, two fusion cassettes were obtained by fusion PCR technique. The first fusion cassette including the upstream fragment of Clf and selectable marker gene was amplified using primer C1F/Hph-R. The second fusion cassette including the downstream fragment of Clf and selectable marker was amplified using primer Hph-F/C2R. In third round PCR, the deletion cassettes was amplified using primer C1F and C2R. For ClUrase gene deletion, the deletion cassettes were generated with C3F/R and C4F/R described as above. These deletion cassette products were introduced into protoplasts of CX-3 described as Liu et al.55. Hygromycin resistant transformants were isolated and selected, and used to detect the hph gene using primers Hphts and Hphta, and the target gene using primer Clf-F and Clf-R by PCR. To generate the RA domain of Clf deletion (Clf ΔRA), a 2.5-kb fragment containing the SAM domain and native promoter, 1.5-kb of Clf amplified with primers SAM-F/R was cloned between BamHI and XbaI sites of pCAMBIA1300th as pCAMBIA1300th-SAM. A 1.3-kb fragment containing protein kinase domain of Clf amplified with primers KI-F/R was cloned between XbaI and HindIII sites of pCAMBIA1300th-SAM as pCAMBIA1300th-SAM-K. The vector pCAMBIA1300th-SAM-K was co-transformed into protoplasts of the Clf gene deletion mutant described as Liu et al.55, followed by screening for Neomycin-resistance. All primers in this study are listed in Table S2.

Analysis of vegetative growth rate, conidiation, conidial and appressorial morphology and infection assays

The analysis of vegetative growth rate and conidiation were as described by Gao et al.9. Conidial morphology was observed using a light microscope. Three independent experiments were performed with 3 replicates each time, and 100 conidia were observed in each replicate. To observe appressorium structure and position, the conidia were inoculated onto artificial cellophane lying flat on PDA medium and examined under an Olympus Microscope BX51 (Olympus, Center Valley, PA, USA) at 6, 9 and 12 h respectively. Images were photographed using Olympus cellSensTM standard software.

For infection assays, conidial suspensions (1 × 106 conidia/ml in 2% glucose and 0.02% Tween 20 solution) or mycelial plugs were used in sprays or inoculated separately on detached leaves. Three independent experiments were performed with five replicates each time.

Additional Information

Accession codes: All genes in this manuscript had been submitted to GenBank. The gene accession numbers are Clg2p (HQ655805), Clf (KT336108), and ClUrase (KT336109).

How to cite this article: Liu, T. et al. Clg2p interacts with Clf and ClUrase to regulate appressorium formation, pathogenicity and conidial morphology in Curvularia lunata. Sci. Rep. 6, 24047; doi: 10.1038/srep24047 (2016).