Characterization of HOG1 homologue, CpMK1, from Cryphonectria parasitica and evidence for hypovirus-mediated perturbation of its phosphorylation in response to hypertonic stress


  • Seung-Moon Park,

    1. Institute for Molecular Biology and Genetics, Basic Science Research Institute, Chonbuk National University, Dukjindong 664–14, Jeonju, Chonbuk 561–756, Korea.
    Search for more papers by this author
  • Eun-Sil Choi,

    1. Institute for Molecular Biology and Genetics, Basic Science Research Institute, Chonbuk National University, Dukjindong 664–14, Jeonju, Chonbuk 561–756, Korea.
    Search for more papers by this author
  • Myoung-Ju Kim,

    1. Institute for Molecular Biology and Genetics, Basic Science Research Institute, Chonbuk National University, Dukjindong 664–14, Jeonju, Chonbuk 561–756, Korea.
    Search for more papers by this author
  • Byeong-Jin Cha,

    1. Department of Agricultural Biology, Chungbuk National University, Cheongju, Chungbuk 361–763, Korea.
    Search for more papers by this author
  • Moon-Sik Yang,

    1. Institute for Molecular Biology and Genetics, Basic Science Research Institute, Chonbuk National University, Dukjindong 664–14, Jeonju, Chonbuk 561–756, Korea.
    Search for more papers by this author
  • Dae-Hyuk Kim

    Corresponding author
    1. Institute for Molecular Biology and Genetics, Basic Science Research Institute, Chonbuk National University, Dukjindong 664–14, Jeonju, Chonbuk 561–756, Korea.
      E-mail; Tel. (+82) 63 270 3440; Fax (+82) 63 270 3345.
    Search for more papers by this author

E-mail; Tel. (+82) 63 270 3440; Fax (+82) 63 270 3345.


We examined the biological function of cpmk1, which encodes a MAPK of Cryphonectria parasitica, and its regulation by mycovirus. Sequence comparisons revealed that cpmk1 had highest homology with osm1, a hog1-homologue from Magnaporthe grisea. A growth defect was observed in the cpmk1-null mutant under hyperosmotic conditions, indicating that cpmk1 functionally belongs to a hog1 subfamily. Immunoblot analyses indicated that the CpMK1 pathway was affected specifically in hyperosmotic conditions by the hypovirus CHV1-EP713. Moreover, the virus-infected hypovirulent UEP1 strain also exhibited severe osmosensitivity compared to the virus-free isogenic strain EP155/2, thus providing additional evidence for viral regulation of cpmk1 in response to a hypertonic stress. Besides osmosensitivity, disruption of cpmk1 resulted in several, but not all, hypovirulence-associated changes, such as reduced pigmentation, conidiation, laccase production and cryparin expression. However, the cpmk1-null mutant exhibited an increased accumulation of pheromone gene transcripts. Virulence assays of the cpmk1-null mutant revealed reduced canker area, but not as severe as that of UEP1. These results suggest that mycoviruses modulate the MAPK and thereby provoke the aberrant expression of target genes, some of which are likely to be implicated in viral symptom development.


The chestnut blight fungus, Cryphonectria parasitica (Murr.) Barr, has been responsible for the virtual disappearance of the chestnut orchards in North America since the early 20th century. However its hypovirus infection is known to cause hypovirulence. Hypovirulence is a good example of biological control (Van Alfen et al., 1975; Anagnostakis, 1982; Nuss, 1992). Strains containing the double-stranded (ds) RNA Cryphonectria hypovirus (CHV) show the characteristic symptoms of hypovirulence, and display hypovirulence-associated changes, such as reduced pigmentation, sporulation, laccase production and oxalate accumulation (Havir and Anagnostakis, 1983; Elliston, 1985; Rigling et al., 1989). Interestingly, the symptoms caused by hypoviral infection appear to be the result of the aberrant expression of specific sets of fungal genes in the hypovirulent strain, which may include the genes for cutinase, laccase, cryparin and mating pheromones (Rigling and Van Alfen, 1991; Choi et al., 1992; Varley et al., 1992; Zhang et al., 1993; 1994). Thus, the chestnut blight fungus C. parasitica and its hypovirus represent a useful model system to study the mechanisms of fungal gene regulation by mycoviruses. Because the phenotypic changes in the fungal host are pleiotropic, albeit coordinated and specific, it has been suggested that the hypovirus disturbs one or several regulatory pathways (Nuss, 1996). Accordingly, several studies have shown the involvement of a signal transduction pathway during viral symptom development.

Several genes that encode components of the various signal transduction pathways in C. parasitica have been cloned and characterized. Included in this group are genes for the heterotrimeric G-proteins and their putative regulator (Choi et al., 1995; Kasahara and Nuss, 1997; Kasahara et al., 2000), and a novel kinase (Kim et al., 2002). In addition, many other genes are under investigation (Dawe and Nuss, 2001). However, the components of the mitogen-activated protein kinase (MAPK) signal transduction pathway have not yet been cloned and characterized.

The mitogen-activated protein kinase (MAPK) signal transduction pathway is utilized by eukaryotic cells to transduce a wide variety of cellular signals through a step-wise phosphorylation relay. This cascade appears to be well-preserved in a variety organisms, ranging from yeast to human (Herskowitz, 1995; Schaeffer and Weber, 1999), and consists of three functionally interlinked protein kinases: MEEK (MAP kinase kinase kinase), MEK (MAP kinase kinase) and MAPK. Several MAPKs have been cloned from various phytopathogenic fungi, including the cereal leaf pathogens Magnaporthe grisea and Cochliobolus heterostrophus (Xu and Hamer, 1996; Xu et al., 1998; Dixon et al., 1999; Lev et al., 1999), the cucumber leaf pathogen Colletotrichum lagenarium (Takano et al. 2000), the maize pathogen Ustilago maydis (Mayorga and Gold, 1999; Muller et al., 1999), the broad-host-range necrotroph Botrytis cinerea (Zheng et al., 2000) and the soil-borne pathogens Fusarium oxysporum and Nectria haematococca (Li et al., 1997; Di Pietro et al., 2001). Studies on the biological functions of the fungal MAPK reveal that it is involved in many different pathways of growth, differentiation and pathogenicity in plant pathogenic fungi (Xu, 2000).

In this study, we cloned the cpmk1 gene for the MAPK of C. parasitica, and provided evidence that it is hypovirus regulated. In addition, we examined the putative biological functions of CpMK1 related to hypovirus-associated traits.


Characteristics of the cpmk1 gene

Using degenerated primers that were derived from the conserved subdomains II and VI (Kultz, 1998), a 300 bp DNA fragment was amplified and sequenced. The closest homologue for this cloned fragment was the osm-2 gene of Neurospora crassa, which shared 82% and 92% identity at the nucleotide and amino acid sequence levels respectively. Southern blot analysis of restriction enzyme-digested C. parasitica genomic DNA indicated that the cloned PCR product was present in the C. parasitica genome as a single-copy gene (data not shown). This PCR product was used to screen a λ DASH genomic library of C. parasitica (Stratagene, La Jolla, CA) and to isolate a clone that contained the entire MAPK gene. Two of the 40 000 plaques screened were positive, and a 6.5 kb BamHI-digested λ clone that contained the C. parasitica MAP kinase gene cpmk1 was selected for further analysis. Based on the genomic sequence, we used RT-PCR to clone the near-full length cDNA fragment using primers at nucleotide positions (nt) − 5–21 and nt 1256–1285 (relative to the start codon). The cpmk1 gene consisted of four exons, with three intervening sequences of 64 bp, 59 bp and 82 bp. The deduced cpmk1 protein product (CpMK1) consisted of 359 amino acids, with an estimated molecular mass of 40.9 kDa and a pI of 5.36 (the GenBank accession number for cpmk1 is AY166687). We found a canonical TATAAA box at nt − 356 in the promoter region of cpmk1. The sequence around the first ATG was in good agreement with Kozak's consensus sequence in that the nt-3 position was the A in CATCATG. The poly(A) signal (AATAAA) was observed 149 bp downstream of the translational stop codon.

Sequence comparisons of the deduced CpMK1 and other MAP kinases revealed the presence of all of the conserved subdomains, including the hallmark dual phosphorylation site TXY (at positions 171–173) (Fig. 1A) (Kultz, 1998). Homology search indicated that the cloned CpMK1 is highly related to the fungal MAP kinases OSM1 from M. grisea (94% identity), HOG1 from S. cerevisiae (82%) and STY1 from S. pombe (80%), and belongs to a subfamily of the yeast stress-activated protein kinase (YSAPK). CpMK1 is only 40% and 32% identical to FUS3 and SLT2 from S. cerevisiae respectively. The phylogenetic analysis indicated that the cloned CpMK1 belonged to the HOG1-related group of MAPKs (Fig. 1B).

Figure 1.

A. Amino acid sequence alignment of the predicted cpmk1 gene product with other MAPKs of the YSAPK subfamily. Identical amino acids are highlighted in white on the black background. The dashes indicate gaps in the alignments. The two closed circles indicate the characteristic residues in the dual phosphorylation site of TXY. OSM1, STY1 and HOG1 are HOG1 homologues of M. grisea, S. pombe and S. cerevisiae respectively.
B. Phylogenetic comparison of fungal MAPKs. A tree is constructed through the upgma (unweighted pair group method using arithmetic means) using Genetyx program (Software Development, Tokyo, Japan). The numbers above the horizontal lines are the frequency with which a given branch appears in 1000 bootstrap replications. The amino acid sequence of MAP kinases from M. grisea (MG), S. cerevisiae (SC), S. pombe (SP), C. albicans (CA) and C. parasitica (CP) were analysed by the growtree algorithm of the GCG package to create the phylogram. The cpmk1 sequence data have been submitted to GenBank under the accession no. AY166687.

To determine the transcriptional regulation of the cpmk1 gene by hypoviruses, total RNA was obtained from Ep155/2 and its isogenic hypovirulent strain UEP1 on days 1, 3, and 5 of culture, and hybridized with a cpmk1 probe. No transcript was detected by Northern blot analysis, even after prolonged exposure times, which suggests that cpmk1 is expressed at very low levels in these two C. parasitica strains (data not shown).

Kinase activity of E. coli-expressed CpMK1

Expression of the full-length CpMK1 in E. coli was examined by SDS-PAGE. Following gentle purification with sonification and nickel-affinity chromatography, a single 50 kDa band was observed for CpMK1. The observed size of the recombinant protein was slightly larger than the calculated mass because of the presence of the hexa-histidine tag. Autophosphorylation of CpMK1 was examined by the addition of [γ-32P]ATP to a reaction buffer that contained only the E. coli-expressed CpMK1. Autophosphorylation of CpMK1 was observed with a prolonged exposure (>3 day), although the enzymatic activity of the E. coli-expressed CpMK1 could not be fully restored because of the lack of phosphorylation by the MAPK kinase. Phosphorylation of a common substrate of MAPKs, myelin basic protein (MBP) (Huang et al., 2000), was also observed, which suggests that the cloned gene encodes a protein with kinase activity and belongs to a subgroup of the MAPK subfamily (Fig. 2).

Figure 2.

Kinase assay of CpMK1 using myelin basic protein.
A. Lane 1 shows the phosphorylation pattern of MBP in the absence of E. coli-expressed CpMK1. Lanes 2, 3, and 4 show the phosphorylation patterns of MBP in the presence of 0.2 ng, 0.4 ng and 0.8 ng, respectively, of E. coli-expressed, full-length CpMK1.
B. Coomassie brilliant blue-stained SDS-PAGE of the samples in A. Arrows indicate the phosphorylated MBP and its derivative. The numbers at left refer to the protein sizes in kDa.

Viral regulation of cpmk1

Immunoblot assay using cell-free extracts of C. parasitica was applied to examine whether the pathway that involved cpmk1 was affected by hypovirus CHV1-EP713 infection. Because the CpMK1 contains residues, MTGYSTR, surrounding the dual phosphorylation site (underlined), which appears to be similar to corresponding residues, MTGYATR, of human p38 MAPK, we used the phospho-p38 MAPK antibody to examine the viral regulation of phosphorylation of CpMK1 after hypertonic induction. As shown in Fig. 3, the phosphorylation of CpMK1 in EP155/2 strain was increased fourfold after hypertonic induction. In hypovirulent UEP1 strain, the phosphorylation level of CpMK1 before hyperosmotic induction was nearly the same as the wild type, but no such increase was detected by the treatment with acute osmotic stress indicating viral specific perturbation of CpMK1 phosphorylation in response to the acute hyperosmotic stress. No phosphorylation band corresponding to CpMK1 was observed in the cpmk1-null mutant (described below), which indicated that phosphoepitope reacted with the phospho-p38 MAPK antibody was specific to the phosphorylated CpMK1. Equal amounts of cell-free extract were confirmed by Bradford assay followed by Coomassie blue staining of a parallel gel.

Figure 3.

Western blot analysis of the phosphorylated-CpMK1 in response to hypertonic induction. The phosphorylation of CpMK1 was monitored using antibody specific for dually phosphorylated p38 MAPK.
A. Autoradiogram of the phosphorylated-CpMK1 after hypertonic induction.
B. Relative band intensity of autoradiogram in A. The band intensity was quantified by densitometry.
C. Coomassie brilliant blue-stained SDS-PAGE of the samples in A. The numbers at the top of each lane indicate the time after hyperosmotic induction, and the strains are designated above the line.

Construction of the cpmk1-null mutant

The cpmk1-null mutant was constructed by site-directed recombination during integrative transformation. A linear PCR fragment that contained the disrupted cpmk1 gene was used to transform the virus-free C. parasitica EP155/2 strain. A total of 120 single-spore transformants were screened by PCR using the inner and outer primers, which corresponded to the 3’-end of the coding region of cpmk1 and the region 56 bp upstream of the 5’-end of cpmk1 in the replacement vector, respectively (Fig. 4A). All but one of the transformants showed the expected 1.3 kb fragment of the wild-type cpmk1 allele, whereas one putative transformant (TdMK1-23) exhibited only the 3.0 kb fragment of the disrupted allele (data not shown). The cpmk1-null mutant was further confirmed by Southern blot analysis (Fig. 4B). As shown in Fig. 4B, TdMK1-23 did not hybridize with the probe from the deleted region, and its hybridization pattern with the 4.0-kb EcoRI-fragment, which contained the 3’-flanking region of cpmk1, differed from that of the wild type, as expected.

Figure 4.

Restriction map and Southern blot analyses of the cpmk1-null mutant (TdMK1-23) and the wild-type EP155/2.
A. Restriction map of the cpmk1 genomic region and the gene-replacement vector pDmk1, which contains 600 bp of the SphI/SalI fragment from the left-flanking region, and 515 bp of the SalI/EcoRI fragment from the right-flanking region. The arrows show the direction of transcription. B, BamHI; E, EcoRI; S, SalI; Sp, SphI.
B. Southern blot analysis of the wild-type EP155/2 strain (lane 1) and the cpmk1-replaced transformant TdMK1-23 (lane 2). All of the DNA samples were digested with EcoRI. The blots were probed with the 4.0 kb EcoRI fragment (probe A) and the 0.6 kb SalI fragment (probe B). The TdMK1-23 transformant has undergone the desired replacement at cpmk1, as evidenced by the changes in the size and number of bands that hybridized with probe A, and the lack of hybridization with probe B. The probes are indicated in the restriction map in (A).

Osmosensitivity of the cpmk1-null mutant

The cpmk1-null mutant did not show any growth defects in standard media, such as PDAmb or EP complete liquid media (Fig. 5A). However, the cpmk1-null mutant showed dramatic growth inhibition on a hypertonic plate that contained 2 M sorbitol, compared with the wild-type EP155/2 strain (Fig. 5B). Similar inhibition patterns of hyphal growth were observed when the cpmk1 mutant was subjected to hyperosmotic stresses, which were independent of the nature of the osmolyte, such as NaCl, MgCl2, and KCl (data not shown). Moreover, growth inhibition under hyperosmotic conditions was also observed in the virus-containing isogenic strain UEP1, which is consistent with the concept that cpmk1 is specifically regulated by the presence of the mycovirus. Thus, we conclude that cpmk1 expression is not required for growth under standard conditions, but is required in situations of hyperosmotic stress and CHV1-EP713 infection.

Figure 5.

Phenotypic characteristics of the cpmk1-null mutant. The numbers 1, 2, 3, and 4 indicate the EP155/2, UEP1, cpmk1-null mutant (TdMK1-23) and cpmk1-complemented strains respectively.
A. Colony morphology on PDAmb.
B. Colony morphology on PDAmb that was supplemented with 2 M sorbitol.

To ensure that the phenotypic changes attributed to TdMK1-23 were due to gene replacement of cpmk1, we complemented TdMK1-23 in trans with a wild-type allele of cpmk1 gene. The resulting benomyl-resistant transformant was used to examine growth and osmosensitivity. Neither growth defect in standard media nor growth inhibition under hypertonic conditions was observed in the cpmk1-complemented transformant (Fig. 5), which confirmed unequivocally that the phenotype changes in the cpmk1-null mutant were due to the disruption of the cpmk1 gene.

Phenotypic changes in the cpmk1-null mutant

The cpmk1-null mutant showed an almost complete absence of pigmentation when grown on the PDAmb medium. The mycelia, which consisted of aerial conidiomata, were maintained as hyaline to subhyaline forms, and rarely became pigmented, even with prolonged cultivation (Fig. 5A). The number of conidia per plate was significantly reduced in the cpmk1-null mutant (4.3 × 107), which produced conidia to the same extent as the virus-containing isogenic strain UEP1 (5.2 × 107). However, no reduction in conidiation was observed for the cpmk1-complemented transformant relative to the wild-type strain EP155/2 (4.1 × 109 versus 4.2 × 109).

The surface hydrophobicity of the mutant was determined by placing a water droplet on the surface of the aerial hyphae. Normally, the surface of hyphae are so hydrophobic that drops of water added to the surface of aerial hyphae remain as discrete drops, whereas the hydrophobin deletion mutants quickly absorb the drops of water. Both the wild type and the TdMK1-23 mutant showed non-absorbed, intact water droplets after long periods of incubation (>12 h), whereas the cryparin-null mutant from the previous study (Kazmierczak et al., 1999) showed the characteristic water-soaked phenotype. However, Northern blot analysis using RNA prepared from liquid culture revealed that the accumulation of cryparin transcripts decreased in the cpmk1-null mutant (Fig. 6A) to a level that was less than that of the hypovirulent strain UEP1. These results may suggest that regulation of cryparin in aerial hyphae is not affected by cpmk1 but expression in submerged culture condition may be under regulation of cpmk1. Therefore, we isolated RNA from mycelial mat on cellophane layered on the top of solid medium and examined the expression of cryparin (Fig. 6B). Similar downregulation of cryparin expression was observed in cpmk1-null mutant, which suggested that the expression of the cryparin gene was cpmk1-dependent and its expression was reduced to a large extent, but not enough to promote the appearance of the water-soaked phenotype. Interestingly, several fungal MAPKs are involved in the regulation of hydrophobin; these MAPKs belong to the yeast and fungal extracellular signal-regulated kinase (YERK1) subfamily (Madhani et al., 1999; Di Pietro et al., 2001) and not to the YSAPK family.

Figure 6.

Molecular characteristics of the cpmk1-null mutant.
A. Northern blot analysis of Crp, Mf2/1 and Gpd. Total RNA from liquid culture was extracted on each of the days listed above the lane. The numbers on the top of lanes refer to days after inoculation.
B. Northern blot analysis of Crp and Gpd. Since the transcription of viral-regulated cryparin gene was affected by growth phase (Zhang et al., 1994), total RNA was extracted from mycelial mat that reached half (H, 4 days after inoculation) or edge (F, 9 days after inoculation) of plate as described in the section of Experimental procedures. The identity of each strain is given above the line. Equal loading of RNA samples was confirmed in a parallel blot that was hybridized with the Gpd probe as an internal control and in an ethidium bromide-stained gel (rRNA).

Tannic acid-inducible laccase expression is also mediated by the CpMK1 pathway

Laccase activity was examined on a plate that contained tannic acid. The TdMK1-23 mutant showed reduced levels of brown coloration, which indicates reduced enzyme activity (Fig. 7). However, no difference in enzyme activity was observed between the wild-type and complemented strains as expected. Growth inhibition of the cpmk1 mutant was not observed on the tannic acid-containing plate, indicating that the reduction of laccase production is independent of growth inhibition and is possibly caused by hypertonic stress. At least three different forms of laccase (the LAC1 and LAC3 extracellular forms, and the LAC2 intracellular form) have been reported in C. parasitica (Kim et al., 1995). The previous study on the disruption of the extracellular laccase gene (Lac1) indicated that an inducible laccase (Lac3) was responsible for enzymatic activity on Bavendamm plates. Therefore, the tannic acid supplement appeared to act as an extracellular signal to induce the corresponding Lac3 expression, which was mediated by the cpmk1-dependent pathway. Moreover, the tannic acid stress stimulus appeared to differ from hypertonic stress in that it did not inhibit the growth of the cpmk1-null mutant, as was obvious on plates that contained 2 M sorbitol. Thus, it is conceivable that cpmk1 responds differently to tannic acid stress than to hyperosmotic response. This is a clear example of a situation in which a given signalling component, CpMK1, is used in more than one pathway within the same cell in response to different signals, in this case, tannic acid and high osmotic stress.

Figure 7.

Colonies were grown on tannic acid-containing medium, as described previously (Rigling et al., 1989). The level of brown coloration correlated with the laccase activity of each strain. The numbers 1, 2, 3, 4, and 5 indicate the EP155/2, ectopic transformant (TMK1), cpmk1-complemented, cpmk1-null mutant (TdMK1-23) and UEP1 strains respectively.

Virulence of the cpmk1-null mutant is reduced, but not to the level of the hypovirulent strain

The pathogenicity test, which was performed on the excised bark of a chestnut tree (Lee et al., 1992), indicated that the cpmk1-null mutant is not as virulent as EP155/2 (Fig. 8). Compared with the wild type, complemented, and hypovirulent strains, the cpmk1-null mutant produced necrotic areas of intermediate size on excised bark, which indicates that although the loss of function of cpmk1 results in a slight decrease in virulence, the virulence of the cpmk1-null mutant remains higher than that of the CHV1-EP713-infected strain.

Figure 8.

Virulence assay using excised tree barks as described previously (Lee et al., 1992). The numbers 1, 2, 3, and 4 indicate the EP155/2, UEP1, cpmk1-null mutant (TdMK1-23), and cpmk1-complemented strains, respectively.
A. A representative figure of excised tree barks 1 week after inoculation.
B. Three replicates of each strain were used and each experiment was repeated three times. The lesion measurement values are shown as the means ± standard deviation (mm2).

Repressive effect of cpmk1 on the mating pheromone gene Mf2/1

Northern blot analysis indicated that accumulation of transcripts from the mating pheromone gene Mf2/1 increased considerably in the cpmk1-null mutant, whereas the expression of Mf2/1 in UEP1 was downregulated, as expected (Fig. 6A). The expression level of Mf2/1 transcripts was sixfold higher than that of the wild type, which underlines the repressive role of the CpMK1 pathway on the regulation of the mating pheromone expression. However, no measurable phenotypic changes, such as altered time of first appearance, number of stromal pustules that contained perithecial necks, female fertility, or ascospore production, were observed. Therefore, further studies using the cpmk1-null mutant will explain consequences of overexpression of the mating pheromone gene related to the mating process.


Hypovirulence is a good model system in which to study how mycoviruses specifically regulate fungal gene expression. One possible explanation for this phenomenon is the modulation of regulatory gene(s), which results in a broad spectrum of phenotypic changes in a virus-specific manner (Kazmierczak et al., 1996; Nuss, 1996).

Many of the MAPKs from plant pathogenic fungi have been cloned, and they are implicated in infection-related morphogenesis and penetration. However, most of the pathogenicity-associated MAPKs are from foliar pathogens and belong to the YERK1 family. Based on sequence similarities, catalytic activity, and response to hypertonic stress, we conclude that the cloned cpmk1 encodes a MAPK that is homologous to the yeast HOG1, which is a representative member of the YSAPK subfamily. Few MAPKs have been studied from fungi that cause disease in hardwoods. We would expect the virulence factors of hardwood pathogens to differ from those of foliar or soil-borne pathogens.

The M. grisea gene for OSM1 is a functional homologue of the yeast HOG1 MAP kinase gene that regulates arabitol synthesis in hyphal cells in response to hypertonic stress (Dixon et al., 1999). The osm1 mutant showed normal growth and development under standard growth conditions. In addition, the osm1 mutant formed the appressoria and exhibited unaltered virulence, which indicates that OSM1 is dispensable for fungal pathogenesis under laboratory conditions. In this study, the cpmk1 gene of C. parasitica was found to encode a functional homologue of the yeast HOG1 MAP kinase gene, and shown to have an essential role in the response to hyperosmotic stress. In addition, the cpmk1-null mutant showed a normal growth rate under standard growth conditions, which indicates that the CpMK1 is not essential for vegetative growth. However, the cpmk1-null mutant showed pleiotrophic phenotypic changes, which induced a reduced level of pigmentation and conidiation, reduced laccase production on tannic acid plate, and a slight reduction in virulence, whereas the osm1 mutant showed only reduced conidiogenesis under standard culture conditions. These differences clearly explain that although the components of the signal transduction pathways may be homologous, the biological function of the identical component may be quite dissimilar, which underlines the danger of relying on homology data to infer specific function.

In addition to phenotypic changes, a molecular symptom of the downregulation of hydrophobin (cryparin) gene transcripts was also observed in the cpmk1-null mutant when grown in liquid and solid media. However, surface hydrophobicity is similar in cpmk1-null mutant and wild type suggesting that the expression of the cryparin gene was reduced to a large extent, but not enough to promote the appearance of the water-soaked phenotype. It has been known for C. parasitica that the cryparin is mainly accumulated in the densely aggregated aerial part of mycelia such as fruiting structures (Carpenter et al., 1992). Thus, it is likely that a small amount of cryparin is sufficient to prevent aerial hyphae from becoming water-soaked. Several fungal MAPKs have been shown to be involved in the regulation of hydrophobin, but these proteins belong to the yeast and fungal extracellular signal-regulated kinase (YERK1) subfamily (Madhani et al., 1999; Di Pietro et al., 2001), rather than the YSAPK subfamily. This indicates that different organisms use different mechanisms to modulate the same target gene or its product. Moreover, overexpression of the mating pheromone gene in the cpmk1-null mutant is of interest. The repressive role of HOG1 and its homologues probably represents a strategy to prevent miscommunication between HOG1 and the other response pathways. Saccharomyces cerevisiae undergoes abnormal activation of the FUS3 pheromone-responsive pathway as a result of mutations in the hog1 gene (Hall et al., 1996; O’Rourke and Herskowitz, 1998). In addition, osm1 disruption in M. grisea led to multiple rounds of appressorium formation (Dixon et al., 1999), which is governed by a functional homologue of FUS3, called PMK1 (Xu and Hamer, 1996). Therefore, examination of the cpmk1-null mutant suggests that CpMK1 also has a repressive role in the prevention of abnormal cross-talk between the hyperosmotic stress pathway and another response pathway, which appears to be the regulating pathway of pheromone gene expression. Unlike the budding yeast in which the pheromone gene expression is under the direct control of the pheromone responsive pathway, in N. crassa pheromone expression is subject to forms of regulation that are independent of the mating pathway itself (Bobrowicz et al., 2002). Therefore, it is likely that the pheromone gene expression in C. parasitica occurs through mechanisms other than the pheromone responsive pathway. Note is worthy to be paid that in U. maydis pheromone gene transcription is regulated by the cAMP pathway (Kruger et al., 1998). Recently, we isolated another C. parasitica MAPK that belongs to the YERK1 subfamily. Therefore, it will be of interest to examine what the biological function of this gene might be and whether CpMK1 has a repressive effect on the expression of this gene.

The cpmk1 mutant exhibited reduced laccase production on tannic acid-containing plates, which suggests that cpmk1 acts in signalling during the interaction of C. parasitica with the chestnut tree in which tannic acid, part of the defence barrier, is abundant. In general, the HOG1 pathway in S. cerevisiae appears to be specific for hyperosmotic stress, whereas the analogous pathway in S. pombe is mediated by the Sty1 MAPK and controls responses to a wider spectrum of environmental stresses, including oxidative stress, UV exposure and heat shock (Brewster et al., 1993; Brewster and Gustin, 1994). In addition to osmosensitivity, laccase production was also drastically reduced in the cpmk1-null mutant. The observation of normal growth on tannic acid-containing plates suggests that the stress imposed by tannic acid differs from that generated by hyperosmosis, which results in severe growth inhibition of the cpmk1-null mutant. These facts indicate that CpMK1, which is the functional homologue of HOG1 in C. parasitica, mediates a cellular response to stress other than hyperosmotic stress and thus partly resembles the Sty1 pathway in fission yeast.

The general functions of the HOG1 subfamily probably include osmoregulation, stress responses and pathogenesis (Xu, 2000). The current study showed phenotypic changes, such as reduced pigmentation, conidiation, laccase production and potential additional functions of HOG1 in pathogenic fungi. These phenotypes provide circumstantial evidence that hypovirus infection and loss-of-function mutant of cpmk1 have similar pleiotropic effects on development of C. parasitica. Recently, we reported a novel protein kinase, cppk1, which was transcriptionally upregulated and related to a subset of symptoms, which included pigmentation, conidiation and mating capability. Thus, it seems likely that symptoms of pigmentation and conidiation are under the control of various pathways, all of which are required for the maintenance of colony morphology. Western analysis using the phospho-specific antibody revealed viral perturbation of CpMK1 phosphorylation in response to the acute hypertonic stress. These results, together with the finding of growth inhibition of the hypovirulent strain under hyperosmotic conditions, strongly suggest that the CpMK1 MAP kinase pathway is affected by the presence of mycovirus, which appears to induce inactivation of CpMK1 pathway by lowering the degree of CpMK1 phosphorylation in response to various environmental stresses.

Experimental procedures

Fungal strains and growth conditions

The CHV1-713-containing hypovirulent C. parasitica strain UEP1 and its isogenic virus-free strain EP155/2 (ATCC 38755) were maintained on potato dextrose agar containing l-methionine (100 mg l−1) and biotin (1 mg l−1) (PDAmb) under constant low-level light at 25°C (Kim et al., 1995). Liquid mycelial cultures were grown in EP complete medium (Puhalla and Anagnostakis, 1971). The methods for preparation of the primary inoculum for liquid cultures and culture conditions were described previously (Kim et al., 1995). Acute hyperosmotic induction was performed as follows: ten 0.5 cm-diameter agar plugs that contained actively growing young hyphae were inoculated on cellophane that was layered on the top of PDAmb medium and incubated until the mycelia reached the end of the Petri plate. The cellophane and the actively growing C. parasitica were then transferred to a plate that contained the appropriate osmolytes and incubated further. The assessment of response to chronic hyperosmotic stress was based on the level of hyphal growth on PDAmb that was supplemented with the appropriate osmolytes.

Cloning and characterization of protein kinase cpmk1

Degenerate primers that were specific for consensus nucleotide  sequences within the most conserved subdomains (II/VI) of the MAPKs were designed (Kultz, 1998). The primer sequences were: MK1-F1 (forward), 5′-GTNGC NATRAARAARAT-3′; MK1-R1 (reverse), 5′-GGYTTNANRTC NCKRTG-3′. The PCR was conducted as described previously (Kim et al., 2002). The 300 bp PCR amplicon was cloned into the pGEMT vector (Promega). The inserted DNA fragments of positive bacterial clones were sequenced using the dideoxynucleotide method, and analysed before being used as a hybridization probes for genomic λ library screening according to the previously described procedure (Sambrook et al., 1989).

In order to obtain the cDNA clone of cpmk1, PCR using reverse transcriptase (RT-PCR) was performed with the cMK1-F1 (forward) 5′-CCATCATGGCTGAATTCGTGCGA GCC-3′ and cMK1-R (reverse) 5′-CTCTATTATTGGCCGTT GAACTCCCCCATG-3′ primers. The cDNA was sequenced using the dideoxynucleotide method and synthetic oligonucleotide primers.

Southern and Northern blot analyses

Genomic DNA from C. parasitica was extracted using the method described by Churchill et al. (1990). DNA (10 µg) was digested with restriction enzymes, blotted onto a nylon membrane, and hybridized with radioactively labelled cpmk1.

RNA from liquid culture was extracted as described previously (Kim et al., 1995). RNA from solid culture was also prepared from mycelial mat grown on cellophane that was overlayered on the top of PDAmb. Plate-grown mycelial mat (0.2 g) was harvested by scraping with a sharp razor and used for RNA extraction at microcentrifuge tube scale according to the method of Choi et al. (1992) with a modification of lysis buffer ( Powell and Van Alfen, 1987). Samples containing 20 µg were used for the examination of gene expression in the aerial hyphae. Northern blot analysis was conducted as described previously (Kim et al., 1995). The levels of cpmk1 transcript were assessed using the glyceraldehyde-3-phosphate dehydrogenase gene (Gpd) as the internal control for gene expression in C. parasitica (Choi and Nuss, 1990).

Heterologous expression of cpmk1 in E. coli

The full-length cpmk1 protein product CpMK1 and the truncated CpMK1 that contained the conserved dual phosphorylation sites (from residues 171–173) were expressed in E. coli as hexahistidine fusion proteins, and purified by nickel-affinity chromatography according to the manufacturer's instructions (Novagen). The cDNA that encoded the full-length CpPK1 was amplified by PCR using the primers: 5′-CACATATGGCTGAATTCGTGCGAGCC-3′ (forward) and 5′- C G CGC GG C CG CT TGG CCGT TGAACT CCCCCATG - 3′ (reverse); the primers were modified to incorporate restriction sites (underlined) for NdeI and NotI respectively. The full-length 1074 bp cpmk1 gene was inserted into the NdeI/NotI sites of the expression vector pET28b. In order to generate the truncated CpMK1, the full-length expression plasmid was digested with XhoI to give a smaller construct that carried residues 1–207 of CpMK1 and had a C-terminal deletion of 151 residues. The resulting recombinant plasmids were transformed into E. coli strain BL21. Induction, purification, and confirmation of the recombinant CpPK1 were conducted using the anti-hexahistidine antibody according to the manufacturer's instructions (Novagen). The E. coli-derived inclusion body was solubilized, and then refolded by step-wise dilution of the denaturants using dialysis (Creighton, 1990).

Kinase activity of CpMK1

The MAPK activity of the E. coli-expressed CpMK1 was assayed by measuring the incorporation of 32P from [γ-32P]-ATP into the myelin basic protein (MBP) (Huang et al., 2000). Refolded CpMK1 was added to the reaction buffer, which contained 5 µg of MBP, 50 mM NaCl, 20 mM Tris-HCl (pH 7.5), 5 mM magnesium acetate, 0.1 mM DTT, 0.1 µM ATP, and 2 µCi [γ-32P]ATP in a total volume of 20 µl (Morawetz et al., 1996). To test for autophosphorylation, the purified CpMK1 was used in a kinase assay without MBP. The kinase assay of cell-free extracts was conducted as described previously (Kim et al., 2002).

In order to examine whether mycovirus affected in a specific manner the signal cascade that involved CpMK1, levels of phosphorylated CpMK1 after hyperosmotic induction were examined by immunoblotting with antibody specific for dually phosphorylated p38 according to the manufacturer's instructions (phospho-p38 MAPK antibody; Cell Signaling Technology). Acute hyperosmosis was induced by transferring cellophane that was layered with actively growing mycelia onto a new plate that contained 2 M sorbitol (Lendenfeld and Kubicek, 1998). The rapid induction of the HOG1-related MAPK pathway in response to osmotic stress was known (Gaits and Russell, 1999; Warmka et al., 2001). Therefore, the cell-free extracts were prepared from cultures 30 min and 60 min after induction. Levels of phosphorylated CpMK1 were quantified by densitometry using GS-800 Calibrated Densitometer (Bio-Rad).

Construction of a replacement vector and fungal transformation

The replacement vector pDmk1, which was designed to favour double-crossover integration events, was constructed as follows: the 1.5 kb SphI/EcoRI fragment that contained the cpmk1 coding region was ligated into SphI/EcoRI-digested pUC19, and the resulting plasmid was digested with SalI to replace the 600 bp cpmk1 coding region with the 2.3 kb SalI fragment of pDH25 (Cullen et al., 1987), which also carried the hygromycin phosphotransferase gene cassette (hph). A linearized fragment that contained the cpmk1 coding region interrupted by the hygromycin phosphotransferase gene cassette was generated by PCR using gene-specific primers, and used to transform the virus-free EP155/2 strain.

The functional complementation of the cpmk1-null mutant using a wild-type allele was carried out. The complementing vector pCmk1 was constructed by insertion of 2.6 kb blunt-ended SalI fragment of pSV50 that contained the benomyl resistant cassette (Orbach et al., 1986) into blunted-ended SphI-digested pCpmk1, which carried a full-length cpmk1 gene. The resulting vector was then used to transform the cpmk1-null mutant.

Protoplast preparation and transformation were performed as described previously (Churchill et al., 1990; Kim et al., 1995). Transformants were selected on top of agar that was supplemented with 150 µg ml−1 hygromycin B (Calbiochem) and with 1.5 µg ml−1 benomyl (DuPont), respectively, passaged 3–4 times on selective media, and single-spore isolated, as described previously (Kim et al., 1995). Polymerase chain reaction and Southern blot analysis was conducted on the genomic DNA of the transformants to check for the replacement and in trans complementation of the cpmk1 gene.

Characterization of the cpmk1–null mutant

The phenotypic and molecular changes of the cpmk1-null mutant were compared with those of the wild-type EP155/2 and hypovirulent UEP1 strains. Phenotypic changes in pigmentation, conidiation and mating capability were measured as described previously (Kim et al., 1995). Virulence test using excised bark of a chestnut tree was conducted according to Lee et al. (1992). Laccase activity was measured by colouring of Bavendamm's medium (0.05% tannic acid, 1.5% malt extract, and 2.0% agar), on which the strains were grown (Rigling et al., 1989). Alterations in the expression of the virus-regulated cryparin (Crp) and mating pheromone (Mf 2/1) genes were examined by Northern blot analysis (Kim et al., 1995; 2002).


This work was supported by a 1999's Korea Research Foundation Grant (99–005-D00070). We thank the Research Center for Industrial Development of BioFood Materials at Chonbuk National University for kindly providing the facilities for this research.