First published online December 22, 2006; 10.1105/tpc.105.037861
The Plant Cell 18:3686-3705 (2006)
© 2006 American Society of Plant Biologists
A Multidrug Resistance Transporter in Magnaporthe Is Required for Host Penetration and for Survival during Oxidative Stress[W]
Chuan Bao Sun,
Angayarkanni Suresh,
Yi Zhen Deng and
Naweed I. Naqvi1
Fungal Patho-Biology Group, Temasek Life Sciences Laboratory and Department of Biological Sciences, National University of Singapore, Singapore 117604
1 To whom correspondence should be addressed. E-mail naweed{at}tll.org.sg; fax 65-6872-7007.
 |
ABSTRACT
|
|---|
In prokaryotes and eukaryotes, multidrug resistance (MDR) transporters use energy-dependent efflux action to regulate the intracellular levels of antibiotic or xenobiotic compounds. Using mutational analysis of ABC3, we define an important role for such MDR-based efflux during the host penetration step of Magnaporthe grisea pathogenesis. Mutants lacking ABC3 were completely nonpathogenic but were surprisingly capable of penetrating thin cellophane membranes to some extent. The inability of abc3 to penetrate the host surface was most likely a consequence of excessive buildup of peroxide and accumulation of an inhibitory metabolite(s) within the mutant appressoria. Treatment with antioxidants partially suppressed the host penetration defects in the abc3 mutant. abc3 was highly sensitive to oxidative stress and was unable to survive the host environment and invasive growth conditions. ABC3 transcript levels were redox-regulated, and on host surfaces, the activation of ABC3 occurred during initial stages of blast disease establishment. An Abc3-green fluorescent protein fusion localized to the plasma membrane in early appressoria (and in penetration hyphae) but became predominantly vacuolar during appressorial maturity. We propose that ABC3 function helps Magnaporthe to cope with cytotoxicity and oxidative stress within the appressoria during early stages of infection-related morphogenesis and likely imparts defense against certain antagonistic and xenobiotic conditions encountered during pathogenic development.
|
INTRODUCTION
|
|---|
Transmembrane proteins belonging to the ubiquitous ATP binding cassette (ABC) superfamily have been identified in several genera representing prokaryotes and eukaryotes (Higgins, 1992 ). Identification and subsequent sequence comparisons of >100 genes encoding ABC transporters revealed that the ABC proteins possess one or two well-conserved nucleotide binding folds of  200 amino acid residues and contain the Walker A and B motifs and the SGG(Q) signature (Michaelis and Berkower, 1995). Several members of this large superfamily function in the transport of cytotoxic agents across biological membranes and help maintain a reduced intracellular level of toxins or metabolites (Driessen et al., 2000). ABC transporters play a major role in the multidrug resistance (MDR) mechanism that operates in mammalian tumor cells. The MDR subfamily of efflux pumps includes the MDR P-glycoproteins. Several MDR-type P-glycoproteins have been shown to catalyze the ATP-dependent efflux of antitumor agents during chemotherapy of cancer cells (Cole et al., 1992; Gottesman and Pastan, 1993). In addition, a number of P-glycoproteins of the MDR family have been implicated in the active extrusion or maintenance of a broad range of compounds, namely, solutes, peptides, hormones, lipids, and drugs (Kolaczkowski et al., 1998).
The ascomycete Magnaporthe grisea causes blast disease in several monocot plant species (Ou, 1985), including rice (Oryza sativa), and represents a model system to study fungal plant interaction (Valent, 1990). To breach the plant surface, the asexual spores or conidia of Magnaporthe elaborate a specialized infection structure termed appressorium. Upon entry into the host cells, the fungus proliferates by forming penetration hyphae and infection hyphae that help in establishing and spreading the disease (reviewed in Hamer and Talbot, 1998; Talbot, 2003).
Phytopathogenic fungi need to adapt to their specific host environment and subsequently overcome the cytotoxic and antifungal compounds or phytoalexins produced by the plant hosts (Kodama et al., 1992; Dixon et al., 1994; Osbourn, 1996). In filamentous fungi, ABC transporter activity likely involved in energy-dependent efflux of fungicides or phytoalexins has been reported in Aspergillus nidulans (Andrade et al., 2000), Aspergillus fumigatus (Tobin et al., 1997), Botrytis cinerea (Vermeulen et al., 2001), Magnaporthe (Urban et al., 1999), Gibberella pulicaris (Fleissner et al., 2002), and Mycosphaerella graminicola (Stergiopoulos et al., 2003; Zwiers et al., 2003). However, none of these molecules belong to the P-glycoprotein MDR subfamily of the ABC transporters.
In this study, we report the identification and characterization of a novel insertional mutant in Magnaporthe that showed a complete loss of pathogenicity toward host plants. The disrupted gene locus, termed ABC3, encodes an MDR protein with extensive homology to the Cluster II.2–type P-glycoprotein transporters, such as MDR1 in humans. We show that loss of Abc3 protein leads to cessation of the infection-related morphogenesis at the host penetration step. Consequently, mutants lacking ABC3 were incapable of breaching host surfaces, and although inefficiently, were still able to penetrate artificial substrates like cellophane. We show that ABC3 itself is regulated in a redox-responsive manner and likely by a plant signal(s). An interesting finding relates to the involvement of ABC3 in regulating the fungal response to oxidative stress, suggesting that Abc3 protein serves as a novel MDR that is necessary for pathogenicity and the ability of the blast fungus to withstand oxidative damage and the host-specific adverse environment.
|
RESULTS
|
|---|
Identification and Isolation of the ABC3 Locus
In an Agrobacterium tumefaciens T-DNA–based forward genetics approach in Magnaporthe, a mutant strain TMT2807 was identified due to its dramatic and total failure to infect the barley (Hordeum vulgare) cultivar Express (Figure 1A
). Further monoconidial and random ascospore analysis–assisted purification of TMT2807 showed that the pathogenesis defect cosegregated with hygromycin resistance, conferred by a single copy integration of the HPH1-containing T-DNA in this strain (Figures 1B and 1C; see Materials for details). Thermal asymmetric interlaced PCR analysis (Liu et al., 1995) revealed that the T-DNA insertion in TMT2807 disrupted the genomic region corresponding to Contig 67 on Supercontig 5.117 (Magnaporthe Genome Database, Release 5, Broad Institute). Further subcloning and sequence analysis showed that the T-DNA insertion in TMT2807 disrupted a region just proximal (231 bp upstream) to exon 1 of open reading frame (ORF) MGG_13762.5 (Figure 1B) and led to a total loss of transcription of this ORF (data not shown). Analysis of the BAC clone 22C21 (identified using the T-DNA insertion flanks from TMT2807 as probes) revealed the presence of the ORF mentioned above as a 6.3-kb HindIII fragment (Figure 1B). We designated this gene as ABC3 because its predicted product showed a high degree of sequence similarity to the ABCs encoded by the genes belonging to the ABC transporter superfamily.
View larger version (31K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 1. Identification and Characterization of the TMT2807 Mutant.
(A) Barley leaf explants were inoculated with conidiospores of TMT2807 strain or wild-type Magnaporthe and disease symptoms assessed after 9 d.
(B) Schematic representation of the annotated ABC3 locus spanning a HindIII-PvuII fragment from Contig 67 on Supercontig 5.117 in Magnaporthe. Closed bars and short open boxes indicate the coding regions and introns, respectively, and are drawn to scale. RB and LB represent the right and left border sequences of T-DNA (open box) integrated in the mutant strain. Opposing arrows demarcate the genomic region deleted in the abc3 strain that was created using flanking homology (dashed lines) based gene replacement with the hygromycin resistance cassette (HPH1). Restriction enzyme sites HindIII (H) and PvuII (P) have been depicted. Bar = 1 kb and also denotes the probe used for DNA gel blot analysis shown in the next panel.
(C) DNA gel blot analysis of the ABC3 mutants and the complemented strain. Genomic DNA from the wild type, abc3, complemented strain (Comp; abc3 carrying an ectopic single-copy integration of the HindIII fragment described in [B] above), and TMT2807 strain was digested with HindIII and probed with the 1-kb ABC3 fragment or the HPH1-specific fragment. The appearance of the 3.2-kb band in the abc3 strain, with the concomitant loss of the wild-type 6.3-kb ABC3 locus, was diagnostic of the correct gene replacement event. Ectopic integration of the rescue construct in the complemented strain resulted in the retention of the 3.2-kb fragment and the restoration of the 6.3-kb band. The presence of a 0.6-kb fragment in the TMT2807 strain is due to an internal HindIII site at the right-border end of the integrated T-DNA in TMT2807. This integron also accounts for the 7-kb fragment (the ABC3 gene disrupted with HPH1 T-DNA cassette) detected in this mutant. DNA gel blot analysis with HPH1 as probe further confirmed the identity of the fragments described earlier. Molecular size markers in kilobase pairs are indicated.
|
|
Characteristics of the ABC3 Gene and Protein
The ABC3 locus spanned nucleotides 124832 to 131097 on Supercontig 177 (Figure 1B). We obtained a full-length ABC3 cDNA by 3' and 5' rapid amplification of the cDNA ends (RACE) and by several RT-PCR fragments representing the overall ABC3 coding sequence. In each instance, DNA sequence analysis was performed on both the strands of at least two independent clones. Subsequent analysis revealed the existence of 13 exons (Figure 1B) in the ABC3 ORF as opposed to the autocall-predicted 11 exons in MGG_13762.5. In addition, three nucleotide changes were uncovered in exon 3 of MGG_13762.5. The complete nucleotide sequence and annotation details for ABC3 have been deposited in GenBank under accession number DQ156556. Typical regulatory elements related to fungal promoter regions preceded the ABC3 ORF. Based on the cDNA sequence, the ABC3 gene was predicted to encode a 1321–amino acid protein (hereafter, Abc3p) composed of two homologous halves, each with six membrane-spanning segments (ABC transmembrane domain) and an ABC ATPase motif. Abc3p thus showed an overall structure and domain organization typical of the Cluster II.2–type (or Transport Commission number 3.A.1.201) P-glycoprotein MDR transporters (Decottignies and Goffeau, 1997; Saier and Paulsen, 2001).
Comparison of the Abc3 protein sequence with related sequences in SWISS-PROT and other protein sequence resources identified several members of the P-glycoprotein family of ABC transporters. A phylogenetic relationship was established using these sequences (Figure 2
). Individual sequence comparisons and phylogenetic analyses based on ClustalW (Thompson et al., 1994; see Supplemental Figure 1 online), MEGA (Kumar et al., 2004), and Phylip 3.6a (Felsenstein, 1989) revealed that Abc3p likely defines a separate family of fungal MDR transporters distinct from the Pmd1 or the ATRC family (Figure 2, arrows). The phylogram also showed that Abc3p had diverged significantly from its most related Ste6 protein in Saccharomyces cerevisiae (Figure 2). However, there was a paralogous transporter (MG09931.4; 56% similarity) encoded in Magnaporthe (Figure 2). Rather surprisingly, neither Abc3p nor MG09931.4 showed any significant similarity (average 7% similarity and 4% identity) to the Abc1 (Urban et al., 1999) or Abc2 transporter (Lee et al., 2005) reported in Magnaporthe. Abc3 showed the highest similarity to SNOG_05968 from Phaeosphaeria (62% similarity and 43% identity) and was found to be closely related to AtrC from A. nidulans (34% identity and 53% similarity), the fission yeast Pmd1 (54% similarity and 35% identity; Nishi et al., 1992), and the budding yeast Ste6 (46% similarity and 28% identity; Ketchum et al., 2001), whereas Abc1 showed a higher level of similarity to the ScPdr5 and CaCDR1 transporters from yeasts (see Supplemental Figures 2 and 3 online).
View larger version (31K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 2. Phylogram of MDR Proteins Related to Abc3p.
ClustalW (Thompson et al., 1994) and Phylip 3.6a (Felsenstein, 1989) assisted dendrogram depicting the most parsimonious phylogenetic relatedness of Magnaporthe Abc3p with the MDR proteins from the indicated genera within eukaryotes. Bootstrapping (500 replicates/iterations) was used in generating the phylogenetic tree using the neighbor-joining algorithm in Phylip 3.6a. Percentage bootstrap support for each clade (when >50%) is indicated below the branch. B-Link web tool on the National Center for Biotechnology Information (NCBI) network was initially used to select 44 MDR hits related to Abc3p.
|
|
Phenotypic Characterization of the abc3 Strain
Using plasmid vector pFGLabcKO in a one-step gene replacement technique, we created an ABC3 deletion mutant (hereafter referred to as abc3) by replacing the entire 4.61-kb coding region of the ABC3 locus (Figures 1B and 1C) with the HPH1 cassette encoding the hygromycin phosphotransferase function. Hygromycin-resistant transformants (HPH1+) derived from the wild-type Guy11 strain carrying single-copy insertion of the replacement cassette were identified, and the desired gene replacement event (abc3::HPH1) was confirmed by DNA gel blotting (Figure 1C, details in the figure legend). For complementation analysis, a full-length genomic copy of ABC3 (Figure 1B; 6.3-kb HindIII fragment) was introduced into the abc3 strain as a single ectopic integron (Figure 1C). At least two independent strains from each background were examined to assess all the vegetative, reproductive, and pathogenesis-related defects reported here.
When inoculated as mycelial plugs or as conidial suspension on complete agar medium, the abc3 mutant grew slower (30% reduction in colony size; P < 0.05) than wild-type Guy11 (Figure 3A
), although no apparent difference in conidiogenesis or the overall colony morphology was uncovered. However, when crossed to TH3 (mat1-1), the abc3 (mat1-2) mutant showed a slight reduction in its sexual reproduction, displaying a decrease in the total number of perithecia produced per sexual cross (Figure 3B). Such defects were not mating-type specific as judged by analyzing the sexual crosses between an abc3 (mat1-1) and wild-type Guy11 (mat1-2) (Figure 3B) and could be attributed to reduced female fertility in the Magnaporthe strains used (Valent et al., 1991). The complemented abc3 strain behaved like the wild type in the ability to form perithecia. This suggested a divergence of function for Abc3p compared with the related Ste6 protein, which is required specifically for the efflux of only the a-factor pheromone in Saccharomyces. These results indicated that in Magnaporthe, the ABC3 function is required for proper vegetative growth of the mycelia but is dispensable for sexual reproduction. Next, we assessed the germination efficiency and growth of the abc3 conidia. Conidia produced by the abc3 strain were normal in quantity, morphology, and germination compared with the wild-type conidia (Figure 3C). Figure 3D shows that upon germination, the abc3 conidia produced appressoria (average diameter 11.2 ± 0.7 µm; n = 3000, P < 0.05) that were marginally larger than those produced by the wild-type strain (10.4 ± 0.6 µm; n = 3000, P < 0.05).
View larger version (93K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 3. Growth Characteristics and Mating in the abc3 Mutant.
(A) Wild-type, abc3, and an abc3 strain complemented with the full-length ABC3 (Comp) were grown on complete medium (CM) for 1 week and photographed. The growth characteristics are representative of three independent assessments (total n = 27 colonies per strain; P < 0.05). Bar = 1 cm.
(B) Abc3 protein is not required for sexual reproduction in Magnaporthe. Perithecia development by the indicated Magnaporthe strains in individual sexual crosses with the tester strain of the opposite mating type were assessed 3 weeks after inoculation. Mean values (±SD) represent the number of perithecial beaks observed per mating mix on oatmeal agar medium. Quantifications represent three independent experiments covering a total of 30 crosses per strain. WT refers to the cross between Guy11 (mat1-2) and TH3 (mat1-1).
(C) Conidia from the wild type or the abc3 strain were allowed to germinate on CM for 16 h and visualized after acid fuchsin staining. Bar = 20 µm.
(D) The wild type or abc3 conidia were inoculated on hydrophobic inductive surface (GelBond membrane) and the morphology of the resultant appressoria assessed after 24 h. Bar = 10 µm.
|
|
ABC3 Is Essential for Magnaporthe Pathogenesis
To assess the role of Abc3p during the host-associated growth and development, we performed infection assays on the seedlings of the rice cultivars CO39 or NIL127 and barley cultivar Express to test the pathogenicity of the abc3 conidia. The abc3 mutant (like the TMT2807 strain, Figure 1A) displayed total loss of pathogenicity and failed to elicit any visible blast symptoms/lesions on the compatible (CO39) or incompatible (NIL127) rice varieties when compared with equivalent spray inoculations of wild-type Guy11 conidia or the conidia from the complemented strain (Figure 4A
). Under the same conditions, the wild type and the complemented strain caused typical spindle-like, gray-centered, and severe blast lesions that merged into one another on the inoculated rice leaves (Figure 4A). We then quantified the host defense–associated hypersensitive reaction (HR) in the challenged seedlings based on real-time RT-PCR methodology to derive the ratio of rice Pr-1 to Actin gene expression as described by Gilbert et al. (2006). As shown in Figure 4A, the abc3 mutant failed to elicit proper HR during the compatible or the incompatible interaction with the host. Barley leaf explants challenged with the wild type or the abc3 mutant for 72 h were costained with 3,3'-diaminobenzidine (DAB) and Trypan blue to visualize the HR symptoms in the challenged tissue. As shown in Figure 4B, the wild-type strain could cause massive damage and cell death and accumulation of reactive oxygen intermediates in the host leaves, whereas the abc3 mutant failed to elicit any visible HR symptoms in the host and was devoid of such cell death and accumulation of reactive oxygen species, except in rare instances (<1%; arrowhead, Figure 4B; discussed later). To further confirm the lack of HR elicitation, we tested the induction of pathogenesis-related protein 5 (Pr-5; Genbank accession number HVU276225) in the barley leaves challenged for 48 h. The ratio of Pr-5 expression to that of a housekeeping function (Actin10-4, Genbank accession number HVU21907) was used to estimate the level of HR induction. In mock inoculations and in barley leaves challenged with the abc3 mutant, this ratio (Pr-5/HvAct10) averaged 1.4 ± 0.3 (over three replicates; P < 0.05), whereas the ratio was consistently higher and averaged 2.6 ± 0.5 (P < 0.005) when estimated in plants infected by the wild-type fungus. The abc3 mutant could still be recovered as viable colony-forming units from the challenged leaves but only up to 30 h after inoculation (data not shown). Based on the above plant infection data, we concluded that Abc3p is required for Magnaporthe pathogenicity, and upon loss of ABC3 function, the blast fungus is incapable of establishing disease and eventually fails to survive inside the host.
Since concentrated suspensions of abc3 conidia failed to elicit any host lesions upon surface inoculation, we decided to test these conidia by two additional means: (1) through inoculation on abraded host leaves and (2) by direct injection of conidia into the leaf nodes or rice leaf sheaths. These assays showed that abc3 mutant is incapable of causing disease-related host damage even when forcibly introduced through wounded tissue (Figure 4C) or injected directly into the host plants (Figure 4C, bottom panels). Appropriate controls, both positive and negative, were included and are shown alongside in these assays (Figure 4C). Even under such conditions that bypass the requirement of appressorium function, the abc3 mutant failed to elicit proper HR as judged by the ratio of Pr-1/Actin gene expression (Figure 4C).
Rather surprisingly, the mutant appressoria were functional on PUDO-193 cellophane (Clergeot et al., 2001) and penetrated it (Figure 4D, bottom panels), albeit with a much lower efficiency (39% ± 2.5% compared with 58% ± 2.1%; P < 0.005) than the wild-type appressoria. The invasive growth within the cellophane appeared to be compromised in the abc3 strain. Taken together, we conclude that the abc3 mutant is incapable of eliciting proper HR or causing blast disease in the host plants but, although inefficiently, is still capable of breaching and invading cellophane membranes.
Genetic Complementation of TMT2807 and the abc3 Mutant
We introduced pFGLr2 (carrying the full-length ABC3 gene) or pBarKS (control) into the abc3 strain and the TMT2807 mutant. Twenty bialaphos-resistant transformants were screened by DNA gel blot analysis in each instance. Two strains that carried single-copy integration of the ABC3 gene at an ectopic site in each background (TMT2807 and abc3) were used to analyze the suppression of the various phenotypic defects observed in the abc3 mutant. The mild reduction in vegetative growth and mating efficiency observed in the abc3 strain (and TMT2807) was completely suppressed upon introduction of the wild-type ABC3 gene (Figures 3A and 3B, Comp), which also restored its ability to penetrate the host and cause blast disease (Figure 4A, COMP). The complemented strain was found to be as virulent as the wild-type isolate when spray-inoculated on rice seedlings (Figure 4A, COMP). By contrast, the vector control could not suppress these abnormalities (data not shown). Since all the defects in the abc3 mutant (and TMT2807) could be completely restored by reintroduction of the wild-type ABC3 allele, we conclude that the phenotypic changes and functional abnormalities observed in the abc3 strain resulted solely from the disruption of ABC3 function.
Infection Structure–Related Defects in abc3
Next, we addressed the question whether the ability to breach the artificial membranes but not the host tissue was due to a decrease in appressorial turgor in the abc3 strain. We therefore estimated the appressorial turgor through the incipient cytorrhysis assays (Howard et al., 1991; de Jong et al., 1997). Such appressorial collapse assays (Table 1
) revealed that compared with the wild type, the abc3 mutant displayed a slightly lower appressorial turgor, although not statistically significant. This reduction in turgor in the mutant appressoria was particularly evident when the external glycerol addition was in the molar range of 3.5 to 4.5. At lower molar concentrations (1 to 3 M), the internal turgor levels were inferred to be equivalent to those estimated in the wild-type appressoria. As judged by light microscopy and by thin-section transmission electron microscopy (TEM), melanin deposition appeared wild type–like in the abc3 appressoria. We isolated TMT2807 and the abc3 strain as totally nonpathogenic mutants incapable of causing blast disease or host damage (Figures 1A and 4). Therefore, we performed detailed microscopic analyses of the infection cycle to determine which stage of the pathogenesis process was compromised in the abc3 strain. Quantitative appressorium formation assays using conidia from the wild type or abc3 revealed that the ability to germinate and form appressoria on barley leaf or onion epidermal strips or rice leaf sheath was not affected in the mutant (Figure 5A
). However, a striking difference was that compared with the wild type, only a negligible percentage of abc3 appressoria (0.7% ± 0.9%; n = 3000, over three replicates) could produce penetration pegs as judged by aniline blue staining for papillary callose deposits and infection hyphae formation (Figures 5A to 5C; 48 h). In the wild type, the ability to form penetration pegs on barley leaves was 83% (± 2.9; n = 3000) at the equivalent 48-h time point. Even upon extended incubation, the failure to penetrate the host surface remained unchanged (Figure 5C, 72 h). Even 96 h after inoculation, a vast majority of the abc3 appressoria (99%) still failed to enter the host and to elicit callose deposition (data not shown). By contrast, the resultant wild-type infectious hyphae within the host (Figures 5B and 5C) were already in their ramifying and proliferating stage in planta.
To further validate our findings above, we used thin-section TEM-based ultrastructural analysis to ascertain whether abc3 appressoria were indeed defective in host penetration. Such TEM analysis confirmed that the majority of the abc3 appressoria failed to penerate the host surface and did not elicit any visible reaction from the plant (Figure 6A
, abc3, 48 h). We could detect futile attempts at penetration in only two appressoria (out of 142 appressoria sectioned) in abc3 (Figure 6A, 72 h). In each instance, the dense papillary callose deposits obstructing the site of failed entry were clearly visible (Figure 6A). The TEM sections for the wild-type appressoria depicted the successful penetration and elaboration of typical fungal infectious hyphae in the host tissue (Figures 6A and 6B, rice sheath assay). The abc3 mutant failed to elaborate any infection hyphae in barley, onion (Allium cepa), or rice (Figure 6C). Based on these results, we conclude that a defect in appressorium-mediated host penetration leads to nonpathogenicity in mutants lacking the ABC3 function. Taken together, we construe that Abc3p plays an extremely important role in the host penetration step of disease establishment and is probably also required for the in planta spreading of the blast fungus.
View larger version (102K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 6. The abc3 Mutant Fails to Penetrate and Colonize the Host.
(A) Barley leaves were challenged with conidia from the wild type or abc3 and processed for thin-section TEM at the indicated time points after inoculation. Near-median TEM sections were selected for assessment and are represented here. Arrows indicate the nonfunctional penetration pegs elaborated by abc3 appressoria, whereas the callose deposits are marked by an asterisk. hcw, host cell wall. Bar = 2 µm, except for the middle right panel where it denotes 1 µm.
(B) The abc3 strain fails to produce infection hyphae. Conidia from the wild type or abc3 were inoculated on barley leaf explants, onion epidermal strips, or rice leaf sheaths (processed for aniline blue staining after 72 h). Arrow indicates the infection hyphae within the leaf sheath tissue.
(C) Respective quantifications of the infection hyphae from three independent experiments in each instance, where values (±SD) are indicated as percentage points.
|
|
In addition, abc3 was incapable of surviving the host environment since viable abc3 cultures could not be recovered from the penetration assays and the wounding assays beyond 30 h after inoculation. Quantitative cell-viability assays using Phloxine B (which accumulates in dead cells) showed that the abc3 mutant was incapable of surviving the environment encountered within the host irrespective of the mode of entry (Figure 7A
; surface inoculation or injection). Appressorial assays on cellophane revealed that viability and invasive growth is significantly compromised in the mutant (Figure 7A, 48 h; P < 0.05). Moreover, addition of barley leaf extract during appressorial assays on cellophane caused a further and marked reduction in the viability of the mutant therein (Figure 7A, 48h+Extr). The wild-type strain remained largely unperturbed under these conditions. We conclude that the ABC3 function plays a critical role during early stages of disease establishment and is most likely required for the in planta viability and survival of the blast fungus.
Abc3p and MDR
Earlier studies have documented the role of MDR P-glycoproteins in the efflux of a diverse range of compounds, such as steroid hormones (Uhr et al., 2002), mating pheromone (Ketchum et al., 2001), drugs and antibiotics (Nishi et al., 1992), and extrusion of ions or solutes (Kimura et al., 2005). To test whether Abc3p serves similar extrusion function(s), we investigated the drug sensitivity of the abc3 strain. We tested the effect of metabolic poisons, antifungal agents, and antibiotics on abc3 strain and wild-type Guy11. As shown in Table 2
, sensitivity to each compound was determined as the minimum inhibitory concentration of that compound required for inhibiting mycelial growth in wild-type Guy11. Among the drugs tested, abc3 strain showed increased sensitivity to valinomycin and actinomycin D. The MDR correlated with the function of an ABC or P-glycoprotein transporter since it could be reversed by the presence of verapamil (Table 2) in Guy11. Our minimum inhibitory concentration assays did not show any difference in the sensitivity of Guy11 and abc3 mutant for the other compounds tested, such as brefeldin A, gramicidin D, leptomycin B, benomyl, and some azole fungicides.
Having confirmed an important efflux-related role of Abc3p during the mycelial growth phase of Magnaporthe, we then assessed whether Abc3p is also required during the pathogenic phase for a similar MDR function. As shown in Figure 7B, the germ tube growth from abc3 conidia was inhibited upon valinomycin treatment, whereas the germ tube growth in Guy11 could withstand an elevated concentration of valinomycin (20 µg/mL). On proper inductive surfaces, the wild-type conidia germinated and formed appressoria in the presence of valinomycin, albeit at a lower frequency: 61 ± 2% versus untreated wild-type samples, where the rate of appressorium differentiation was found to be 93% ± 5% (Figure 7C). The appressoria formed under these conditions were also substantially smaller. By contrast, the same concentration of valinomycin had a very severe and dramatic effect on the conidia of the abc3 mutant: a complete blockage of conidial germ tube emergence (Figure 7C). It has been proposed that valinomycin can also affect the homeostasis of potassium ions across biological membranes (Daniele and Holian, 1976). However, exogenous addition of potassium chloride did not suppress the effect of valinomycin toward the conidia and appressoria of neither the wild type nor the abc3. We conclude that the Abc3 protein serves an essential multidrug resistance function both during the vegetative and the pathogenic phases of the rice blast fungus and suggest that such MDR activity is most likely directed toward compounds related to valinomycin in their structure and/or properties. Abc3p could also functionally replace its ortholog, Pmd1p, from fission yeast (Figure 7D), signifying the evolutionary conservation and importance of such MDR phenomenon.
Abc3p and Sensitivity to Cellular Stress
Since addition of potassium could not reverse the effect of valinomycin, we reasoned that the Abc3 protein functions through some as yet unknown mechanism to modulate its MDR activity toward such antifungal compounds. We therefore investigated whether Abc3p is required for modulating osmotic, oxidative, and/or related cellular stress conditions. Compared with the wild-type strain, abc3 mutant did not show a difference in its mycelial growth in high concentrations of osmolytes, such as sorbitol or sucrose, or sodium chloride. By contrast, the abc3 mutant was found to be highly sensitive to oxidative stress, and even a small dose (2 mM) of hydrogen peroxide was found to be lethal to the mutant strain (Figure 8A
), whereas the wild-type Guy11 strain remained unperturbed under such adverse growth conditions (Figure 8A). The abc3 mutant also showed similar sensitivity to paraquat and menadione, which stimulate accumulation of reactive oxygen species/intermediates (ROS; data not shown). The sensitivity of the abc3 mutant to such oxidative stress was common to the vegetative and pathogenic growth phases (Figure 8B). Nitroblue tetrazolium–assisted visualization of the accumulation of reactive oxygen radicals in the wild type and the abc3 colonies showed that the mutant hyphae accumulated relatively higher amounts of peroxide and likely a higher amount of ROS (Figure 8C; the dark centers in the wild-type colony are due to the enhanced contrast of the image). Such increased accumulation of peroxides/DAB-positive material was also detected in the abc3 mutant appressoria on artificial membranes and barley leaf explants (Figure 8D, bottom panels). Fluorimetric evaluation of hydrogen peroxide levels in the wild type and abc3 on barley leaves, after staining with the dye chloromethyl-2',7'-dichlorofluorescein diacetate (CM-DCFDA), revealed that the mutant appressoria accrue 3.5-fold higher amounts of ROS/peroxide than those estimated in the wild type (Figure 8E). An earlier study has shown that oxidative stress (accumulation of DAB-positive material) precedes the elicitation of cell death in the interacting epidermal cells and the underlying mesophyll cells in Brachypodium infected with Magnaporthe (Routledge et al., 2004).
View larger version (95K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 8. abc3 Is Highly Sensitive to Oxidative Stress and Accumulates Excess Peroxide and Reactive Oxygen Intermediates.
(A) and (B) Mycelial plugs (A) or equivalent serial dilutions of conidiospore suspensions (B) from the abc3 strain or the wild type were cultured on minimal agar medium containing the indicated amounts of hydrogen peroxide. The results were documented after an incubation period of 1 week.
(C) The abc3 strain accumulates excess peroxide. Three- (top panels) or six-day-old (bottom panels) wild-type or abc3 colonies were stained with nitroblue tetrazolium solution to detect peroxide accumulation.
(D) Peroxide accumulation during pathogenic phase. In vitro and in vivo DAB reactions. Wild-type and abc3 appressoria formed on artificial membranes (top panels) or on barley leaf explants (bottom panels) were stained with DAB to detect the accumulation of hydrogen peroxide (black arrows). White arrows indicate the excess DAB-reactive deposits within the appressorial lumen. Bars = 10 µm.
(E) Total extracts from the wild type or the abc3 appressoria were treated with CM-DCFDA and processed for fluorimetric estimations. Values in the graph represent fold change (mean ± SD; from three independent experiments) in the CM-DCFDA–reactive material from mutant appressorial extracts compared with those in the wild type. Average value for the CM-DCFDA estimates in the wild-type extracts was set at 1.
|
|
To understand whether the pathogenicity defects in abc3 are in any way related to the excess buildup of H2O2 therein, we tried to reduce the levels of peroxide oxidants during appressorium development using several agents, such as 15 µm diphenyleneiodonium,10 µm nordihydroguaiaretic acid, 5 µm rotenone (Chiarugi et al., 2003), or antioxidants (ascorbate or N-acetylcysteine). The addition of diphenyleneiodonium could partially (in 5.1% ± 1.1% appressoria; n = 3000) suppress the host penetration defects associated with the abc3 appressoria (Figure 9A
). However, antioxidant treatment with either ascorbate or N-acetylcysteine restored penetration function in 23% ± 2.5% and 19% ± 1.6% appressoria, respectively (Figure 9A; n = 3000, P < 0.005), as judged by papillary callose deposition. However, the resultant mutant penetration pegs still failed to elaborate proper infection hyphae and were unable to advance the invasion further (data not shown). We conclude that reduction of the intracellular peroxide levels (or oxidative stress) in abc3 leads to a significant suppression of the appressorial defects in this mutant.
In a converse experiment, we tested the effect of exogenous hydrogen peroxide, menadione, or paraquat on the ability of the wild-type appressoria to penetrate the host surfaces. As expected, in the absence of exogenous peroxide, the wild-type appressoria behaved normally and were able to successfully penetrate the host surface as judged by papillary callose deposits upon aniline blue staining. In this instance, the percentage of appressoria that elaborated penetration pegs was 68% ± 2.3% (n = 3000 apppressoria, over three replicates), after the 48-h time point. Addition of 3 mM peroxide to the wild-type conidia caused a significant decrease in host penetration with only 36% ± 2.8% appressoria (Figure 9B; n = 3000, over three replicates) able to breach the host surface. Addition of paraquat or menadione had a slightly reduced effect on host penetration of the wild type compared with treatment with peroxide (Figure 9B). Such oxidative stress caused a significant reduction in the efficiency of appressorium formation (Figure 9B), and peroxide levels exceeding 5 mM blocked germ tube emergence in the wild-type conidia (data not shown). A verapamil-based general downregulation of MDR activity during pathogenic development showed a significant decrease (maximal 74%) in penetration peg formation but without affecting the germ tube emergence or growth. On the basis of these results, we conclude that excessive peroxide levels have an adverse effect on the host penetration capacity of Magnaporthe and construe that reducing the intracellular levels of peroxide (or oxidative stress) in abc3 significantly suppresses the appressorial defects observed in this mutant. Taken together, we conclude that Abc3p performs an important MDR function during pathogenic phase and possibly protects the blast fungus from peroxide-based internal oxidative stress during vegetative and pathogenic development.
abc3 Appressoria Accumulate Toxic Metabolite(s)
Since Abc3p was predicted to be an MDR protein, it raised an attractive possibility that the abc3 appressoria possibly retain cytotoxic molecule(s) that block the appressorial function of elaborating penetration pegs. To address this issue, we prepared intracellular and extracellular extracts (see Methods for details) from the appressoria of the wild type or the abc3 strain and used these extracts individually during plant infections with the wild-type conidia. The said extracts were added to the wild-type conidia in barley leaf infection assays at 0 or 23 h after inoculation. A solvent control at the equivalent concentration was included. The intracellular extract from the abc3 appressoria, when added at the start of the wild-type infection, caused a dramatic (74%; P < 0.005) reduction in penetration peg formation, as judged by aniline blue staining for papillary callose deposits (Figure 10A
, 0 h). There was no decrease whatsoever in the appressorium formation capability under these conditions. The corresponding extract from the wild-type appressoria resulted only in a slight decrease (24%) in host penetration. This decrease by the wild-type extract was comparable to the reduction observed when either the wild type or the mutant extract was added after the infection had proceeded for 23 h (Figure 10A, 23 h). The control extracts did not cause any significant decrease in the host penetration capability of the wild-type appressoria (Figure 10A) and neither did the extracellular extracts from the wild type or the abc3 appressoria (data not shown). These results lead us to hypothesize that the abc3 mutant is unable to efflux certain toxic or inhibitory metabolites, whose accumulation in the mutant appressoria likely inhibits host penetration.
Next, we tested the ability of the intracellular extract from abc3 appressoria to suppress penetration of artificial substrates by the wild type or the mutant. We therefore performed wild-type and abc3 appressorial assays on PUDO-193 cellophane in the presence or absence of this extract. As shown in Figure 10B, addition of the intracellular abc3 extract caused a significant reduction (85% in the wild type and 89% in abc3; P < 0.005) in appressorium function on cellophane membranes. Surprisingly, the intracellular abc3 extract showed minimal effect on the viability of the wild type or the abc3 mycelia (Figure 10C). Interestingly, when included during blast infection assays, the intracellular abc3 extract caused a dosage-dependent and noticeable reduction in the elaboration of disease symptoms on rice leaves (Figure 10D). The said mutant extract also caused an elicitation of visible HR-like symptoms as judged by the occurrence of brown lesions (in the absence of fungal biomass) in treated samples and inferred by the measurement of the ratio of rice Pr-1 to Actin gene expression (Figure 10D; P < 0.05). We conclude that Abc3p is important for the efflux of certain appressorial metabolites, whose accretion has a negative effect on host penetration.
Regulation of ABC3 Expression
Given the above findings that showed the involvement of Abc3p in regulating oxidative stress within the fungal cellular structures, we decided to test whether such stress conditions directly influence ABC3 expression. To this end, we performed a qualitative and quantitative study of ABC3 expression during the infection-related developmental stages of two independent transformants, each carrying a single-copy destabilized green fluorescent protein (deGFP) reporter driven by the ABC3 promoter. As shown in Figure 11A
, the destabilized GFP fluorescence increased dramatically upon treatment with hydrogen peroxide. The fluorescence intensification was found to be fourfold higher compared with that assessed in the untreated control samples (data not shown) and was maximal after a 9-h treatment (Figure 11A). Low levels of paraquat or valinomycin elicited a similar increase in the ABC3 promoter-driven destabilized GFP expression (data not shown) during appressorium development. Based on these results, we concluded that the upstream regulatory elements of the ABC3 gene respond to the model molecules that generate reactive oxygen radicals and that the ABC3 transcript is likely regulated in a redox-responsive manner during infection-related development in Magnaporthe.
View larger version (76K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 11. Redox-Responsive and Host-Dependent Regulation of ABC3 Expression.
(A) Conidia from the ABC3(p):deGFP strain were allowed to undergo appressorium development for 24 h in the absence (Control) or presence (Treated) of 1 mM H2O2. Epifluorescent microscopic observations were performed at the indicated time points to detect destabilized GFP expression driven by the ABC3 promoter. Bar = 10 µm.
(B) Host surface–dependent regulation of ABC3 transcript levels. Semiquantitative RT-PCR–derived products amplified using ABC3-, Mg CHAP1-, or Mg TUB1-specific primers from total RNA extracted from the wild-type strain grown for the indicated hours after inoculation (hpi) on barley leaves. Negative control (-RT) refers to the RNA sample being processed without a reverse transcriptase step prior to the PCR amplification. GD refers to the PCR amplification of the respective fragments from the genomic DNA samples using the corresponding primer sets. Molecular mass standards (in kilobase pairs) are shown.
|
|
To study whether ABC3 expression is also governed by host signals, we performed semiquantitative RT-PCR (Soundararajan et al., 2004) analysis of total RNA extracted at various time intervals from barley leaves challenged with wild-type conidia. Along with ABC3, we included TUB1 (ß-tubulin) and Mg CHAP1 as controls. CHAP1 transcript has been shown to be under redox regulation in Cochliobolus (Lev et al., 2005). As shown in Figure 11B, the highly abundant TUB1 showed hardly any difference in its expression during the duration of the assay. ABC3 expression was always weaker compared with TUB1 but showed an induction at the 18- to 20-h stage after inoculation. Its expression remained stable and strong after the induction at this time point and continued to be so up to 36 h after inoculation (Figure 11B). CHAP1 expression also peaked at 12 to 18 h but declined slightly after the 24-h time point. Taken together, we conclude that the ABC3 promoter responds positively to the redox status of the cell and that the ABC3 gene expression is likely influenced by the host signals too, with an induction just before the fungus readies itself to enter the host tissue.
Subcellular Localization of GFP-Tagged Abc3p
The predicted secondary structure of Abc3p suggested that it is an integral membrane protein with 12 transmembrane helices. To determine the distribution pattern of Abc3p within the various cell types in Magnaporthe, the ABC3-GFP fusion construct was introduced into the Guy11 strain. Two transformants carrying a single copy of the ABC3:GFP allele tagged at the genomic locus and two control strains were identified by requisite PCR analyses and further confirmed by DNA gel blot and protein gel blot analyses (data not shown). GFP detection was performed by either epifluorescence microscopy or using confocal laser scanning fluorescence microscopy. Compared with the wild type, strains expressing Abc3-GFP showed no obvious difference in growth or pathogenesis, suggesting that Abc3-GFP is fully functional. Abc3-GFP was adjudged to be a plasma membrane resident protein based on the distinct green fluorescence in the mycelial outer membranes (Figure 12A
). No detectable GFP signals were observed in the conidia of the ABC3:GFP strain (Figure 12A), although the germ tubes showed a faint GFP signal in the plasma membrane (data not shown). A time-course analysis during appressorium development and maturity suggested that the Abc3-GFP protein was highly expressed and concentrated in the appressorial plasma membrane right from the incipient stage onwards (Figure 12B). However, at a later stage (17 h after germination) during the appressorium maturation, Abc3-GFP signal was predominantly found in the vacuolar compartments as well (Figure 12B, 20 h), although the plasma membrane residency was diminished substantially. Colocalization studies to confirm vacuolar residency involved staining with LysoTracker Red DND-99 or with Neutral Red (Figure 12C, vacuoles). Continuing further, the Abc3-GFP localization during host penetration and invasive growth revealed that Abc3-GFP localized to the plasma membrane of the newly formed penetration hyphae (Figure 12D) but was undetectable in the resultant infection hyphae. Consistent with its role in appressorium function and its localization pattern described above, we conclude that Magnaporthe Abc3p is a transmembrane protein found in the mycelia and predominantly in the appressoria, where we speculate, it is most likely required for the efflux of toxic metabolite(s) and/or for regulating the oxidative stress within these highly important fungal infection structures.
View larger version (67K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Figure 12. Subcellular Distribution of Abc3-GFP Fusion Protein.
(A) Mycelia and conidia harvested from the strain expressing an ABC3-GFP fusion protein were imaged using laser scanning confocal microscopy to detect the enhanced GFP signal. Bars = 10 µM.
(B) Appressorium development in the germ tubes of the ABC3-GFP strain was monitored over a 24-h period. At the indicated time points after germination, GFP epifluorescence was imaged to locate the Abc3-GFP fusion protein. Bar = 10 mm.
(C) Vacuolar distribution of Abc3-GFP. Appressoria formed by the Abc3-GFP strain were stained with LysoTracker Red DND-99 (left panel, vacuoles) or with Neutral Red (right panels) at the 20-h time point to visualize vacuoles. DIC, differential interference contrast. Bars = 10 µm.
(D) Confocal microscopy–assisted analysis of Abc3-GFP localization in the penetration structures (arrow) elaborated by the appressoria in PUDO-193 membrane. Bar = 10 µm.
|
|
|
DISCUSSION
|
|---|
Isolation and Molecular Analysis of the TMT2807 Mutant
We isolated TMT2807 as a nonpathogenic mutant in Magnaporthe. Molecular genetics and further characterization of TMT2807 helped identify the MDR P-glycoprotein encoding gene ABC3 as a novel pathogenicity factor in Magnaporthe. To confirm the role of Abc3p in Magnaporthe pathogenesis, we created and characterized abc3 mutants in two separate wild-type backgrounds, namely, Guy11 and B157. Such studies clearly demonstrated that loss of Abc3p in Magnaporthe leads to a complete loss of pathogenicity toward rice and barley. The abc3 colonies were slow growing compared with the wild type but were largely unaffected in morphology and in conidia formation. Genetic complementation analyses confirmed that the defects seen in TMT2807 and abc3 were due solely to the loss of ABC3 function in these mutants.
ABC3 Function Is Important for Host Penetration
We have shown here that ABC3 is required for Magnaporthe pathogenesis and that its function is most likely critical at the host penetration step during blast disease establishment. The vast majority (99%) of abc3 appressoria failed to breach the host surface even upon extended incubation. As judged by papillary callose deposits and TEM analyses, only a negligible number of mutant appressoria managed to make futile attempts at host penetration but were able to elicit weak HR within the host tissue during compatible and incompatible interactions as well as upon host inoculation through wounds or by injection. We feel that the abc3-host interface still needs to be defined further in this regard and will certainly gain from recent studies that used lesion mimic mutants (Park et al., 2004) or resistant cultivars (Gilbert et al., 2006) to assess the interaction of rice with Magnaporthe mutants that are compromised for host penetration.
Cytorrhysis assays (Howard et al., 1991; de Jong et al., 1997) revealed that there was no significant reduction in the overall turgor generated by the abc3 appressoria, despite their average size being slightly larger than the wild-type appressoria. Interestingly, the abc3 appressoria were capable of penetrating cellophane membranes, albeit less efficiently than the wild-type strain. However, subsequent invasive growth and spread within the cellophane membranes were significantly compromised in the abc3 mutant, which also showed reduced viability under these conditions.
Accumulation of Inhibitory Metabolite(s) in the abc3 Mutant
Our results indicated that the Abc3 transporter activity is most probably required for the timely efflux of certain inhibitory metabolite(s) during the host penetration step of blast infections. Intracellular extracts from abc3 appressoria, when applied exogenously, caused a striking reduction in the wild-type strain's ability to penetrate host surfaces and cellophane membranes, thus suggesting that the abc3 mutant is defective in the efflux of toxic metabolite(s) whose accumulation in appressoria likely inhibits penetration.
The ability of abc3 to penetrate cellophane (although inefficient) is still puzzling given that the inhibitory metabolites discussed above were made from mutant appressoria on GelBond polyester films. We speculate that it is probably the amount of the inhibitory metabolite(s) and/or the timing of their production that are the limiting factors that allow the mutant to escape the maximal activity of these inhibitors on cellophane. Another likely possibility is that maximum turgor generation is probably not necessary on such surfaces. It remains to be seen whether the host surface also plays a role in the extrusion and/or the efficacy of these inhibitory metabolite(s). In preliminary experiments, exogenous addition of the toxic metabolite(s) during wild-type infection assays caused a discernable reduction in blast symptoms and elicited noticeable HR in the host leaf even in the absence of the blast fungus. Further experiments are certainly necessary to really ascertain the true nature of these HR-like symptoms. Future experiments will be directed at identifying the biochemical nature and function of these inhibitory metabolites trapped in the abc3 appressoria.
Novel Functions for Abc3-Mediated Efflux in Magnaporthe
Abc3 protein showed extensive similarity to several MDR-like proteins from filamentous fungi, most notably Phaeosphaeria, Fusarium, Neurospora, and Aspergillus species. Intragenome BLAST (Magnaporthe Genome Database, Broad Institute) searches using Abc3p as query revealed that the Magnaporthe genome encodes at least 76 ABC-like transporters and further helped us identify a paralog (MG09931.4; 56% similarity) within the Magnaporthe proteome. Such paralogs were also uncovered in Aspergillus, Fusarium, and Neurospora species, and their occurrence was suggestive of a recent gene duplication event. Phylogenetic analyses revealed that Abc3p most likely defines a distinct class of MDR transporters that is separate from the ones represented by Pmd1 and AtrC. Abc3p belongs to the MDR P-glycoprotein subfamily of ABC transporters that modulate the efflux of a broad range of compounds, such as sugars, inorganic ions, heavy metal ions, peptides, lipids, metabolic poisons, and drugs (Kolaczkowski et al., 1998 |
TOP
ABSTRACT
INTRODUCTION
