Introduction

Hyphal tip growth is a key feature of filamentous fungi. To support the tip growth, a specialized apparatus is required, which contains five major components: the Spitzenkorper, the polarisome, the exocyst, recycling endocytosis, and Cdc42 GTPase module (Harris 2009; Jones and Sudbery 2010; Kohli et al. 2008; Lichius et al. 2012). In the budding yeast Saccharomyces cerevisiae, the polarisome was identified as a 12S multi-protein complex that is consist of Spa2, Pea2, and Bud6/Aip3, and interacts with Bni1 (Sheu et al. 1998). Besides, it has been observed that Msb3 and Msb4 bind to the N-terminal of Spa2, suggesting that these two proteins are also components of the polarisome (Tcheperegine et al. 2005). Spa2 is a scaffold protein physically interacting with all other polarisome components through specific binding domains (Sheu et al. 1998; van Drogen and Peter 2002). Spa2 is involved in the direction and control of cell division, and localizes to the site of cell growth (Gehrung and Snyder 1990; Roemer et al. 1998). Bud6 is an actin-binding protein, which has an additional function in microtubule plus-end capture at the cell cortex, with contributions of formins (Amberg et al. 1997; Delgehyr et al. 2008). Pea2 was shown to be important for the stability and localization of Spa2 (Valtz and Herskowitz 1996).

SPA2 orthologous genes have been identified in a number of ascomycetous fungi, and this protein family seems to be conserved. In Ashbya gossypii, AgSPA2 is required for area determination of growth at hyphal tip (Knechtle et al. 2003). In Aspergillus nidulans, Spa2 homologs localize to hyphal tips, where they regulate hyphal morphology and the rate of hyphal extension (Virag and Harris 2006), and the Aspergillus niger SpaA was suggested to play a role in ensuring maximal polar growth rate (Meyer et al. 2008). In Candida albicans, CaSPA2 was found to be important for virulence (Zheng et al. 2003). However, in the basidiomyceteous pathogen Ustilago maydis, SPA2 was dispensable for pathogenicity (Carbo and Perez-Martin 2008).

In this study, we isolated and functionally characterized an SPA2 orthologous gene in the rice blast fungus Magnaporthe oryzae. Deletion of MoSPA2 led to slow vegetative hyphal growth and unusual branching of vegetative hyphae. MoSPA2 is also important to maintain normal size of conidia. In the strain expressing the MoSpa2-GFP fusion, accumulated green fluorescence signals as a spot could be observed at vegetative hyphal tips, septa of hyphae and conidial tip cells where germ tubes are usually produced but not in appressoria, infection hyphae and at the septa of conidia. Furthermore, MoSPA2 was found to be dispensable for plant infection.

Materials and methods

Strains and growth conditions

The wild-type strain P131 and its various transformants generated in this study (Table 1) were cultured at 25 °C on oatmeal tomato agar (OTA) plates (Peng and Shishiyama 1988; Yang et al. 2010). Mycelia collected from 2-day-old cultures in complete medium shaken at 150 rpm were used for isolation of fungal genomic DNA and protoplasts. Protoplasts were isolated and transformed with the PEG/CaCl2 method (Park et al. 2004). Media were supplemented with 250 µg/ml hygromycin B (Roche, USA) or 400 µg/ml neomycin (Ameresco, USA) to select hygromycin-resistant or neomycin-resistant transformants, respectively. Genetic crosses were performed as described by Talbot et al. (1996).

Table 1 Fungal strains used in this study
Full size table

Nucleic acid manipulation

Genomic DNAs were extracted from vegetative hyphae with the CTAB protocol (Xu and Hamer 1996). Standard protocols of molecular biology were followed for plasmid isolation, DNA gel blot analyses, and DNA enzymatic digestion (Sambrook and Russell 2001). Probes were labeled with the random primer labeling kit (TaKaRa, Dalian, China). Plasmid constructs were sequenced with an ABI 3730 sequencer (SunBio Company, China). Reverse transcription assays were performed with the PrimeScript RT-PCR Kit (TaKaRa, Dalian, China). All primers (Table 2) were synthesized by Sangon (Shanghai, China).

Table 2 PCR primers used in this study
Full size table

Identification of MoSPA2

A restriction enzyme-mediated DNA integration (REMI) mutant X6149 was generated by transforming the plasmid pUCATPH into the wild-type strain P131 (Lu et al. 1994; Yang et al. 2010). Flanking sequences of the insertion site were isolated using the plasmid rescue approach (Sweigard et al. 1998) and was determined with an ABI 3730 sequencer (SunBio Company, China). BLASTn and BLASTp were used to search for DNA and protein sequences in the GenBank nr database. Protein sequence comparisons and alignment analysis were performed with ClustalX (Saitou and Nei 1987), ESPript (Gouet et al. 1999), and MEGA4 (Tamura et al. 2007).

Complementation assay

The MoSPA2 complementation vector pCSPA2 was constructed by cloning MoSPA2 along with its 1.3-kb upstream sequence and 0.5-kb downstream sequence, which was amplified with the primer pair hbf/hbr (Table 2) into pKN (Yang et al. 2010). After linearization with DraI, pCSPA2 was introduced into the mutant X6149. The resulting transformants were confirmed to harbor a single copy of pCSPA2 by DNA gel blot.

Generation of the MoSPA2 deletion mutants

For generating the MoSPA2 gene replacement vector pKSPA2, the 1.74-kb upstream and 0.83-kb downstream sequences of MoSPA2 were amplified with primer pairs upf/upr and downf/downr, respectively. The resulting PCR products were cloned into the SpeI–PstI and ClaI–KpnI sites of pKNH, respectively (Yang et al. 2010). After linearization with NotI, pKSPA2 was transformed into P131. Hygromycin-resistant transformants were isolated and assayed for resistance to neomycin. The transformants were further screened by PCR with the primer pair inf/inr and out/hphup. The putative Mospa2 null mutants were further confirmed by DNA gel blot.

Generation of the MoSPA2-eGFP fusion

The MoSPA2–eGFP fusion vector pKGSPA2 was generated by cloning MoSPA2 along with its 1.2-kb upstream sequence, which was amplified with the primer pair spa2gfpf/spa2gfpr (Table 2) into pKNTG (Yang et al. 2010). After linearization, pKGSPA2 was transformed into the ∆Mospa2 null mutant.

Plant infection

A 4-week-old seedlings of the rice cultivar ‘LTH’ and 8-day-old seedlings of barley cultivar ‘E9’ were used for infection assays. For the infection assay, conidial suspensions with a concentration of 5 × 104 conidia/ml were sprayed on to the plant leaves, and the inoculated plants were incubated as previously described (Peng and Shishiyama 1988). Lesions were examined and photographed 5–7 days post-inoculation (dpi).

Microscopy

Conidia were inoculated on to barley epidermis cells, and the infection process was observed at 24 and 48 h post-inoculation (hpi). Nuclei and septa were visualized using Hoechst 33258 (Sigma) at 1 μg/ml and Calcofluor White (Sigma) at 10 μg/ml, respectively (Du et al. 2013; Xu and Hamer 1996). Samples were observed and photographed under a Nikon Ni90 fluorescence microscope (Nikon, Japan).

Results

Isolation of the mutant X6149

To identify genes that are important for asexual development and pathogenesis in M. oryzae, we screened the REMI library established in our laboratory (Yang et al. 2010). One mutant, X6149, was isolated after screening over 8,000 transformants and found to be exhibited impaired growth. When incubated on OTA plates for 5 days, the average size of the wild-type colony was 40 mm in diameter, whereas X6149 was only 23 mm (Fig. 1a). To determine whether the defects of X6149 were caused by an REMI event, a genetic cross was made between X6149 (MAT 1-1) and a wild-type strain S1528 with the opposite mating type (MAT 1-2). Among the 34 progeny isolated, all of the 19 hygromycin-susceptible progeny exhibited the wild-type phenotype, and the remaining 15 hygromycin-resistant progeny showed the same defects in colony growth as X6149. These data indicated that the defect in colony growth of X6149 was co-segregated with the hygromycin gene and there was only a single integration site of the exogenous DNA in the genome of this mutant. DNA gel blot hybridization probed with the transforming vector pUCATPH revealed that the mutant X6149 had one 21.0-kb BglII hybridizing band and two 5.1- and 4.5-kb HindIII hybridizing bands (Fig. 1b), indicating that there was only a single integration site of the exogenous DNA in the genome of this mutant, and the exogenous DNA should have two copies.

Fig. 1
figure 1

MoSPA2 was isolated from an REMI mutant X6149. a Colonies of the wild-type P131, the mutant X6149, and one complementation transformant CM20 cultured on OTA plates at 5 dpi. b DNA gel blot analysis of genomic DNA of X6149 and P131 digested with HindIII or BglII and probed with the hygromycin phosphotransferase (hph) gene cassette. The estimated sizes of each band are labeled at right (kb). c Schematic diagram showing the exogenic plasmid pUCATPH integrated into MoSPA2 (MGG_03703.6). Rectangles indicate exons of MoSPA2, ATG represents the putative translation start site, and the triangle indicates the integration site of pUCATPH

Full size image

Identification of MoSPA2

From the mutant X6149, a 13.0-kb plasmid pXB13 with 3.0-kb chromosomal DNA flanking the pUCATPH insertion was isolated. Further sequence analyses showed that the insertion occurred at the second exon of an open reading frame, MGG_03703.6 (Fig. 1c). The deduced protein encoded by this gene shared 50 % identity with Spa2p (GenBank Access No. NP_013079) of S. cerevisiae, and hence was named as MoSPA2. To confirm that MoSPA2 was responsible for the defects of X6149, we carried out a complementation assay by introducing the vector pCSPA2 into X6149. Twenty-three complementation transformants were obtained and several of them, including CM20, were confirmed to harbor a single copy of pCSPA2 by DNA gel blot (data not shown). The colony diameter of CM20 (40 mm) was completely restored to the wild-type level (40 mm) at 5 days post-inoculation (dpi) on OTA plates (Fig. 1a). These results indicated that MoSPA2 was responsible for the defects in vegetative hyphal growth of X6149.

The 3,005-bp open reading frame of MoSPA2 is interrupted by two introns, whose positions were verified by sequencing of the corresponding cDNA. The deduced 951 amino acid sequence has a predicted molecular mass of 103.96 kDa with a pI of 5.21 and shows high degree of overall homology with the Spa2 homologous proteins from other Ascomycetes (Fig. S1). When comparing MoSpa2 with S. cerevisiae Spa2, a common domain named SHD-I (Roemer et al. 1998) could be identified. The amino acid sequence of MoSpa2 SHD-I domain exhibited 46 % similarity with other Ascomycetes sequences. Remarkably, the Spa2 direct repeat (SDR, A(R/Q)xKLxxLxxx(R/Q)Fx(D/E)LxxDVxxELxRR; Roemer et al. 1998) located in the SHD-I domain seems to be extremely conserved from yeast to filamentous fungi (Fig. S2), suggesting that the SDR plays crucial role for protein function.

Generation of the deletion mutants of MoSPA2

To further define functions of MoSPA2, we generated a gene replacement vector pKSPA2 by replacing the MoSPA2 ORF with a hygromycin phosphotransferase gene cassette (hph) (Fig. 2a). Two putative MoSPA2 deletion mutants, MH1 and MH2, were verified by PCR and confirmed by DNA gel blot (Fig. 2b). The 2.1-kb band amplified with the primer pair out/hphup was detected only in the putative Mospa2 null mutants, while a 1.2-kb band amplified with primer pair inf/inr was detected only in the wild-type P131 (data not shown). When probed with a 0.83-kb fragment amplified from pKSPA2 with primer pair downf/downr, the putative MoSPA2 deletion mutant had one hybridizing 6.5-kb HindIII fragment instead of the hybridizing 2.0-kb fragment of the wild-type P131. The MoSPA2 transcript was also detected by RT-PCR in the wild-type P131 and complementation strain CM20, but not in the knockout mutants (Fig. 2c).

Fig. 2
figure 2

The MoSPA2 knockout vector and gene replacement mutants. a The MoSPA2 knockout vector was constructed by amplifying the upstream and downstream flanking sequences with primer pair upf/upr and downf/downr, respectively, and ligated with the hph cassette. H, HindIII. b DNA gel blot analysis of HindIII-digested genomic DNAs of the wild-type P131 and two ∆Mospa2 null mutants MH1 and MH2 hybridized with Probe 1. The estimated size of each band is labeled at right (kb). c RT-PCR was used to confirm the loss of MoSPA2 transcripts from the ∆Mospa2 null mutant MH1. Actin was used as a control

Full size image

Deletion of MoSPA2 affects hyphal growth and branching

To determine the role of MoSPA2 on vegetative hyphal growth, the ∆Mospa2 null mutant, the wild-type P131, and one complementation transformant CM20 were incubated on to OTA plates for 5 days. While the wild-type strain and the complementation transformant formed a fluffy and grayish colony, the ∆Mospa2 null mutant formed a smaller darker colony (Fig. 3a). The average diameter of the wild type and the complementation transformant colonies was approximately 40 mm at 5 dpi. Under the same conditions, the diameter of the ∆Mospa2 colonies was only 27 mm (Fig. 3b). When stained with Calcofluor White, the vegetative hyphae of the ∆Mospa2 null mutant exhibited significantly more septa than those of the wild type and the complementation transformant (Fig. 3c). The average length of sub-apical hyphal cells in the ∆Mospa2 null mutants was 49.4 μm, which was significantly shorter than the 84.5 μm of the wild type and the 84.6 μm of the complementation transformant (Fig. 3d). In addition, the average length of secondary apical hyphal cells in the ∆Mospa2 null mutants was 51.3 μm, which was also significantly shorter than the 85.4 μm of the wild type and the 83.9 μm of the complementation transformant (Fig. 3d). When co-stained with Hoechst 33258 and Calcofluor White, one nucleus was detected in each cell compartment of the vegetative hyphae of the wild-type strain, the complementation transformant and the ∆Mospa2 null mutant (Fig. S3). The ∆Mospa2 null mutant also displayed a lower radial colony growth rate and produced more apical and sub-apical branches than the wild-type P131 and the complementation transformant CM20 (Fig. 3e). These results indicate that MoSPA2 plays a role in polarity maintenance and is required for maximal polar growth rate.

Fig. 3
figure 3

MoSPA2 is required for vegetative hyphal growth. a Colony of the wild-type P131 and the ∆Mospa2 null mutant MH1 cultured on OTA plates at 5 dpi. b The colony diameter of the strains at 5 dpi. The data represent the mean values of a quadruple experiment with a standard deviation. Asterisks indicate a significant difference in the colony diameter between the ∆Mospa2 null mutant and the wild-type strain (or complementation strains) at p = 0.01 according to t test. c Vegetative hyphae of P131, MH1, and CM20 stained with Calcoflour White. White arrows point to the septa. Bar 20 μm. d The average length of the sub-apical cells and secondary-apical hyphal cells of vegetative hyphae in P131, MH1, and CM20. Means and standard deviations were calculated based on three independent experiments by measuring at least 50 mycelia in each replicate. Asterisks indicate a significant difference in the cell length between the ∆Mospa2 null mutant and the wild-type strain (or complementation strains) at p = 0.01 according to t test. e The branch pattern of vegetative hyphae in P131, MH1, and CM20 on cover glass at 2 dpi. Bar 10 μm

Full size image

MoSPA2 is important for conidium morphology

We stained the conidia with Calcofluor White to observe whether disruption of MoSPA2 had any effects on conidium morphology. The ∆Mospa2 null mutant was found to produce smaller conidia (Fig. 4a), and about 60 % of the ∆Mospa2 conidia were smaller than that of the wild type (Fig. 4b). On average, conidia from the wild type and the complementation transformant were 24.5 and 24.6 μm in length, respectively. In contrast, the ∆Mospa2 conidia were approximately 20.4 μm in length (Fig. 4c). In comparison with the wild-type conidia, the lengths of the basal and apical cells of the ∆Mospa2 conidia were significantly reduced than the middle ones (Fig. 4c). Therefore, MoSPA2 is required for maintaining normal conidium morphology.

Fig. 4
figure 4

MoSPA2 is required for maintaining conidium morphology. a Conidia of the wild-type P131, the ∆Mospa2 null mutant MH1, and one complementation strain CM20 stained with Calcofluor White. White asterisks in central panel point to abnormal conidia. Bar 20 μm. b The percentage of conidia with normal size in P131, MH1, and CM20. Asterisks indicate a significant difference between the percentage of conidia with normal size of the mutant and wild-type strain (or complementation strain) at p = 0.01 according to t test. c Average length of the basal, middle, and apical cell of the conidia by P131, MH1, and CM20 and the total length of the conidia. Means and standard deviations were calculated based on three independent experiments by measuring at least 100 conidia in each replicate. Asterisks indicate a significant difference of the cell length of conidia between the mutant and the wild-type strain (or complementation strain) at p = 0.05 according to t test

Full size image

The expression pattern of MoSPA2

To investigate the expression pattern of MoSPA2, a 1.2-kb fragment of the MoSPA2 promoter was amplified and cloned into the vector pKNTG. The resulting plasmid pKGSPA2 containing the enhanced green fluorescent protein (eGFP) was introduced into the ∆Mospa2 null mutant. One of the resulted transformants, CG12, which carried a single copy insertion of pKGSPA2 and restored the wild-type colony growth and conidial morphology, was selected for analyzing sub-cellular localization and expression pattern of MoSpa2. The green fluorescence signals were easily observed in the cytoplasm of conidia and appressoria (Fig. 5a), but was obscurely observed in infection hyphae. Interestingly, an obvious aggregation of green fluorescence signals was observed at the tip of apical or basal cell of conidium where germ tube was produced (Fig. 5a). We also examined the sub-cellular localization of MoSpa2-GFP in vegetative hyphae. In all of more than 300 vegetative hyphae, a strong aggregation of green fluorescence signals was observed at each hyphal tip (Fig. 5b). Moreover, strong green fluorescence signals were observed at the septum of vegetative hyphae but not of conidia (Fig. 5b).

Fig. 5
figure 5

The expression pattern of MoSPA2. a Conidia, appressoria (12 hpi), and infection hyphae (24 and 36 hpi) of a complementation transformant of the ∆Mospa2 null mutant MH1 expressing the MoSpa2–eGFP fusion were examined under differential interference contrast (DIC) and epifluorescence (GFP) microscopy. Bar 10 μm. b Distribution of the MoSpa2-eGFP fusion at the tip of vegetative hyphae. Bar 10 μm

Full size image

MoSPA2 is dispensable for pathogenicity

To investigate the roles of MoSPA2 on plant infection, conidial suspensions of the ∆Mospa2 null mutant MH1, the complementation transformant CM20, and the wild-type P131 were sprayed on to rice seedlings (Fig. 6a). Plants inoculated with the ∆Mospa2 null mutant MH1 formed numerous typical disease lesions as the complementation transformant CM20 and the wild-type P131 did. Similar results were observed with barley seedlings (Fig. 6b). Moreover, conidia produced by the ∆MoSpa2 null mutants were normal in appressorium formation and infection hyphae development on barley epidermal cells (Fig. 6c–e). Thus, MoSPA2 was not required for plant infection.

Fig. 6
figure 6

MoSPA2 is dispensable for plant infection. Seedlings of rice (a) and barley (b) were sprayed with conidia of the wild-type P131, the ∆Mospa2 null mutant MH1, and one complementation strain CM20. Disease lesions were examined and photographed at 5–7 dpi. c Appressorium formation and infection hyphae developments of P131, MH1, and CM20 on barley epidermal cells at 24 and 48 hpi. Bar 20 µm. d The appressoria formation rates of P131, MH1, and CM20 at 24 hpi. Means and standard deviations were calculated based on three independent experiments by measuring at least 100 germinated conidia in each replicate. e The infection hyphae formation rates of P131, MH1, and CM20 at 48 hpi. Means and standard deviations were calculated based on three independent experiments by measuring at least 100 appressoria in each replicate

Full size image

Discussion

In this study, we identified and functionally characterized MoSPA2, a gene encoding a putative spindle pole antigen in M. oryzae and plays important roles in vegetative hyphal growth and conidium morphology. By monitoring the sub-cellular localization of the MoSpa2-GFP fusion in living cells, we could observe an obvious accumulation of the green fluorescence signals located at the tip of vegetative hyphae, which is coincident with its important roles in hyphal growth. Consistently, as a component of the polarisome, Spa2 has been identified from many fungi, including the ascomycete A. gossypii, A. nidulans, A. niger, N. crassa, and the basidiomycete U. maydis, and all of them were shown to localize at the apex of hyphae and was important to hyphal growth (Carbo and Perez-Martin 2008; Crampin et al. 2005; Knechtle et al. 2003; Meyer et al. 2008; Virag and Harris 2006). We also generated the gene deletion mutants of FG00932 in Fusarium graminearum, which was the ortholog of SPA2, and exhibited slow hyphal growth and small colony (data not shown). Therefore, the apex localization of Spa2 is essential to cell polarity and hyphal tip growth. Previously, Spa2 was also found to involve in the direction and control of cell division (Gehrung and Snyder 1990; Roemer et al. 1998). When combined with the observations that no differences were observed in nucleus number of each cell compartment of the vegetative hyphae between the wild-type strain and the ∆Mospa2 null mutant, we speculated that the smaller sizes of the ∆Mospa2 ones should be partially attributed by a shorter cell cycle with the deletion of MoSPA2.

Different with the previous findings that most majority of Spa2 was concentrated within the apex of hyphal tips, most of the MoSpa2-GFP fusion in M. oryzae was also exhibited dotted distribution among the internal compartment of the sub-apical hyphae and its distribution pattern was acropetal enrichment. This phenomenon might partially explain the reasons on cell polarity growth and smaller size of hyphal cells of the ∆Mospa2 null mutants. Like the emergence of germ tubes from a spore, formation of hyphal branches needs the establishment of a new polarity axis. In some cases, such as C. albicans, A. gossypii, lateral branching occurs at sites of pre-existing septa (Harris 2008). In other cases, branching sites may be selected at random, such as N. crassa, and lateral branching generally occurs in the middle hyphal compartment (Harris 2006). M. oryzae preferentially forms lateral branching at sites of pre-existing septa, and the MoSpa2–GFP fusions were shown to be septum localized, which might be responsible for more branches of the colony in the strains with the deletion of MoSPA2. These observations indicated that MoSpa2 can serve as a marker of polarization and is essential for hyphal branching.

The MoSpa2-GFP fusion was also observed in conidia, and especially a bright green fluorescent spot was observed to localize at the tip of apical or basal cell of conidium where germ tube is produced (Patkar et al. 2010). When combined with the observations that the germination tube usually formed from the cell that had a bright green fluorescent spot of the MoSpa2–GFP fusion, MoSpa2 was supposed to ensure polarity development of conidia despite no obvious differences between the ∆Mospa2 null mutant and the wild type on conidial germination pattern (apical- or basal-cell germination; data not shown). However, cell lengths of the ∆Mospa2 conidia were significantly shorter than the wild-type ones, especially the apical cell, which should be attributed by loss function of MoSPA2. During the process of infection-related morphogenesis, no spotted accumulation of the MoSpa2–GFP fusion were observed in appressoria, penetration pegs, and infection hyphae, although weak green fluorescence signals were also rarely detected in them. Consistently, the ∆Mospa2 null mutant had no defects in appressorium formation, penetration peg formation, infection hyphae development, and also virulence toward host plants. Similar results were also found in U. maydis (Carbo and Perez-Martin 2008). Therefore, spotted accumulation of the MoSpa2–GFP fusion might reflect the essential functions of MoSPA2 during the asexual development and conidium morphology, and MoSpa2 has the roles not only in polarity maintenance, but also during early events of polarity establishment. This study also provided solid evidence that there is no direct relevance between vegetative hyphal growth and invasion hyphal development by large scale analysis of pathogenicity genes in the rice blast fungus (Jeon et al. 2007).