DISCUSSION

Despite the fact that several fungal pathogens during early development has been studied by shotgun-based quantitative proteomics analysis (Cagas et al. 2011; Oh et al. 2010; Suh et al. 2012), no study has been conducted on the fungal pathogen, Foc TR4, that causes the most important lethal disease of banana, resulting in substantial losses to the worldwide banana crop (Li et al. 2012). In this study, we conducted the experiments using Complete Medium to culture Foc TR4 for reliably controlling distinct growth stages in mimic of its early development periods in banana host upon infection. With shotgun proteomics analysis, we screened some important candidate proteins/pathways critical to the early growth of Foc TR4 as an initial step, which may also provide insight into subsequent study on pathogenicity mechanisms in Foc TR4 infection to banana host. The four specific stages for early developmental proteomes were intensively investigated (Fig. S1) and the resulting differentially expressed proteins were further statistically and bioinformatically analyzed, and validated by both western blotting and growth inhibition assay by selective fungicides. The significances of this work with identified candidate proteins/pathways for subsequent development of transgenic banana resistant to Foc TR4 and new fungicides are discussed below. In addition, the complex and highly dynamic Foc TR4 proteome along with stage-specific expression patterns for early stages of development reported here provides a benchmark allowing direct comparison of Foc TR4 with other well studied fungi, such as A. fumigatus, A. nidulans etc. in understanding molecular mechanisms regulating spore development (Cagas et al. 2011; Oh et al. 2010; Suh et al. 2012).

Identification and Functional Classification of Differentially Expressed Proteins

A total of 3,035 proteins were confidently quantified during early developmental stages, which represents 18.5% of the predicted Foc TR4 proteome, of which 1,009 from three developmental stages were differentially expressed (>2.0-fold changes against time 0 sample). Majority of these differential proteins were found in 6 biological function groups including carbohydrate and energy metabolism, cell wall biosynthesis and modification, protein biosynthesis, lipid metabolism (ergosterol synthesis), signal transduction and nucleic acids biosynthesis. Interestingly, a recent report showed that the anti-pathogenic (including fungal pathogens) mechanism of most recently developed fungicides has also been involved in the targets from the above 6 categories of biological processes, along with inhibition of mitosis and cell division, induction of pathogen-resistant plants and the use of multiple targets (Turner 2012; Yang C et al. 2011). Thus, we believe that some of the differentially expressed proteins found in this study can be additional candidate targets. Towards this end and given the important findings for Foc TR4 early development, we discuss in details in the following sections on those differential proteins according to their functional categories.

To induce the biosynthesis leading to growth and morphological development, the spore must provide enzymes for energy metabolism. We identified a total of 92 differentially expressed proteins belonging to the carbohydrate metabolism and energy production (Supplementary Table S5) category. Nine succinate-semialdehyde dehydrogenases and aldehyde dehydrogenases were down-regulated (Table 2). This suggests that aldehyde is an important class of intermediate in two-carbon metabolism during early development of Foc TR4. Subsequently, the converted acetyl-CoA would enter the tricarboxylic acid (TCA) cycle for energy generation. Furthermore, several key enzymes of the glyoxylate cycle such as isocitrate lyase and malate synthase showed down in one or all three time points. The glyoxylate cycle is an anaplerotic pathway of the TCA cycle that allows growth on C2 compounds by bypassing the CO₂-generating steps of the TCA cycle (Dunn et al. 2009). Several ATP synthases F₁F₀ (ATP4, ATP5, ATP7, and ATP12), which are part of the catalytic core related to ATP biosynthesis were found to be consistently up-regulated in 3 h, 7 h, 11 h during early development. ATP plays a significant role in free-energy transduction in living cells (Wang et al. 2013). As a result, aldehyde dehydrogenases, ATP synthases and their pathways along with key members of the TCA cycle should be critical to the early development of Foc TR4.

Protein synthesis has been shown to be required for Foc TR4 early development. We identified 85 differentially-expressed proteins involved in translation including 45 distinct ribosome proteins based on KOG classification and GO annotation (Supplementary Table S5). Remarkably, most of the those proteins including 44 distinct ribosome proteins were up-regulated at all three time points, particularly at 3 h and 7 h. This finding is consistent with the expected increase in translation necessary for growth and indicates that proteins involved in translation are dynamically active during the Foc TR4 early development. Our iTRAQ data also indicated that there are 21 differentially expressed proteins involved in amino acids metabolism. One of them is NADP-specific glutamate dehydrogenase (GDH, FOIG_02886T0) which increased about 15 folds, but the NAD-specific GDH (FOIG_04147T0) decreased by 4 folds (Table 2). This is consistent with finding in the yeast Saccharomyces cerevisiae, where two NADP-dependent GDHs encoded by GDH1 and GDH3 catalyze the synthesis of glutamate from ammonium and α-ketoglutarate, while the GDH2-encoded NAD-dependent GDH catalyzes the reverse (DeLuna et al. 2001). Also a number of proteins involved in regulatory functions, protein folding or modification during Foc TR4 early development were identified, including 5 down-regulated thioredoxins and a few down-regulated ATP-dependent Clp proteases, mitochondrial chaperonin, STI1, multifunctional chaperone and other heat shock proteins (Table 2 & Supplementary Table S5).

The fungal cell wall is a complex structure composed typically of chitin, glucans, glycoproteins and other polymers, which changes constantly during cell division, growth and morphogenesis (Adams 2004; Leng et al. 2008). The presence of the 1,3-beta-glucan synthase component FKS1 (FOIG_14431T0), chitin synthases (FOIG_06735T0, FOIG_06738T1), fasciclin and related adhesion glycoproteins (FOIG_06630T0) showed up-regulation, yet the trehalose phosphorylase (FOIG_04571T1), and mannose-6-phosphate isomerase (FOIG_04596T0, FOIG_12838T0) were down-regulated in conidia (Table 2). The results suggest that cell wall synthesis and remodeling are very active during early development in Foc TR4.

Signal transduction is not only an integral component of the host recognition process, but also a regulator for the fungi development and vegetative growth processes. During Foc TR4 early developmental stages, we identified 51 differentially expressed proteins that were associated with signaling transduction (Supplementary S5). Both GTP binding protein and serine/threonine protein kinase involved in cell cycle control were up-regulated in the early development. While majority of signaling transduction proteins involved in Ras-related GTPase, cAMP-PKA and calcium/calmodulin signaling pathways were significantly decreased. Leng et al. (2008) studied the proteome profiles on conidia germination in T. rubrum and found that the Ras-related GTPase and cAMP-PKA pathways may play roles in conidial germination. Although several signaling transduction proteins were found at the early developmental stages of Foc TR4, which is in agreement with the previous reports in other fungal strains, the regulatory mechanism of signal transduction needs to be studied further for Foc TR4.

RNA molecules play a central part in cell metabolism, either by directing the synthesis of proteins or by functioning as essential components of protein synthesis machinery. In this study, 15 differentially expressed proteins were categorized into ATP-dependent RNA helicases, and all of them were up-regulated at 3 h except the dhx8 (Supplementary Table S5). RNA helicases are ubiquitous, highly conserved enzymes that participate in nearly all aspects of RNA metabolism. These proteins bind or remodel RNA or form RNA-protein complexes in an ATP-dependent manner (Jankowsky 2011). The increased activity of ATP-dependent RNA helicases during the early development of Foc TR 4 at 3 h suggests extremely active biochemical metabolisms and physiological states in the fungal cells.

Our experimental data including initial discovery proteomics and verification analyses by western blot, quantitative RT-PCR and Foc TR4 growth inhibition assay of several fungicides, indicate that the ergosterol biosynthesis pathway plays an important role in early development of Foc TR4. Therefore, the following discussion is selectively focused on the results associated with the ergosterol biosynthesis pathway only, as well as the significances and prospects of our findings for subsequent development of transgenic banana resistant to Foc TR4 and new fungicides.

New Target Sites in Ergosterol Biosynthesis Pathway for Inhibition of Foc TR4 Early Growth

For the past two decades, there have been no or few options developed for effectively controlling Fusarium wilt of banana caused by Foc TR4 in field. That is because the soil borne pathogen Foc TR4 can survive in soil for 30 years and prevent from effective uptake of various nutrients and the fungicides once it infects banana (Li et al. 2012). Thus, even the most effective fungicides including Prochloraz yield excellent inhibition of fungi in vitro, none of them appears to practically yield satisfactory results. Therefore development of genetically improved Foc TR4-resistant cultivars is a better choice in the field by banana breeding biologists than creating new fungicides. Previous biochemical and genetic evidence demonstrates that lipid metabolism and secondary metabolism are essential for spore germination during early developmental stages (Christensen and Kolomiets 2011). Out of 84 differentially expressed proteins identified being relevant to lipid metabolism, 7 of them belong to super-pathway of ergosterol biosynthesis (ERG4, ERG6, ERG11/CYP51, ERG13, ERG25, ERG26, HMG1) as shown in Fig. 4A. All 7 proteins were up-regulated at 3 h, 7 h and 11 h (Table 2). This pathway is fungus-specific, and the ergosterol is required for generation of a major constituent of the fungal plasma membrane (Parks and Casey 1995). The ergosterol biosynthesis is via the mevalonate pathway, in which the ERG13 (HMG-CoA synthase) and HMG1 (HMG-CoA reductase) catalyze two key consecutive steps for the conversion of acetoacetyl-CoA to HMG-CoA, and HMG-CoA to mevalonate, respectively (Oh et al. 2008). The ERG11/CYP51 (Cytochrome P450 lanosterol C-14α-demethylase) is an essential enzyme in the biosynthesis of sterols, critical components of cell membranes of all eukaryotic organisms required for the regulation of membrane fluidity and permeability (Lepesheva and Waterman 2007). For many sterols identified in fungi, ergosterol is the most common one required for fungal growth (Fan et al. 2013; Weete et al. 2010). The ERG4, ERG6, ERG25 and ERG26 are also involved in individual specific steps for ergosterol biosynthesis (Fig. 4A). Comparison of the 4 enzymes (ERG6, ERG11/CYP51, ERG13 and ERG25) in ergosterol synthesis by iTRAQ profiling and quantitative RT-PCR (Fig. 5) for all three early development time points indicate that the 4 proteins are consistently up-regulated. In yeast, mutation of ERG11/CYP51, ERG13 and ERG25 led to reduction of function and decreased competitive fitness and even death. Mutation of ERG6 also caused the decreased spore germination, higher susceptibility to stress resistance, as well as reduced vacuolar transport capability (http://www.yeastgenome.org/). Recent studies demonstrated that using HIGS technique, the transgenic Arabidopsis and barley with a recombinant dsRNA of cytochrome P450 lanosterol C-14α-demethylase genes (ERG11/CYP51) from F. graminearum, a fungal plant pathogen causing a devastating disease of wheat and barley, became completely immune to F. graminearum (Koch et al. 2013). Furthermore, in axenic cultures of F. graminearum, in vitro feeding of CYP3RNA, a 791-nt dsRNA complementary to CYP51A, CYP51B, and CYP51C, resulted in effective growth inhibition with half-maximal inhibitory concentration (IC₅₀) of 1.2 nM, as well as altered fungal morphology, similar to those observed after treatment with the azole fungicide tebuconazole, whose action target is also the CYP51 enzyme (Koch et al. 2013). Ghag et al. (2014) found that dsRNA-mediated gene silencing of two vital fungal genes: velvet and ftf1 (Fusarium transcription factor 1) from Foc Race 1 enabled the original susceptible banana cv. Rasthali (AAB subgroup) to develop efficient resistance against Foc Race. Given the fact that both F. graminearum and Foc TR4 belong to the same category of Fusarium and HIGS technology has proven quite successful in producing transgenic plants against F. graminearum and Foc Race1, it is promising to develop transgenic banana with sustainable resistance against Foc TR4 Fusarium wilt by HIGS. Quantitative proteomics results plus multiple verification results based on western blot, RT-PCR and in vitro anti-fungi assays indicate that ergosterol synthesis appears extremely important in the Foc TR4 early development. For ERG6, ERG11/CYP51, ERG13 and ERG25 enzymes, consistent up-regulation was observed at both transcriptional and translational levels for all time points, suggesting their transcripts may have significant effect on the final protein abundance. Additionally, by blast analysis of existing genomic data, no orthologous genes/proteins in banana were found for all up-regulated proteins including ERG6, ERG11/CYP51, ERG13 and ERG25 of ergosterol synthesis pathway. As a result, those are likely to be the ideal candidate targets in subsequent development of new transgenic lines by HIGS toward effective resistance against Foc TR4. It should be noted that despite the fact that ERG4, ERG26 and HMG1 in the ergosterol synthesis pathway were found to be down-regulated by RT-PCR, and showed a negative relationship to their protein abundance, these could still be useful targets for HIGS-based new banana recombinants against Foc TR4. This is because we are confident with our proteomics data in which ERG4 was further validated by western blot (Fig. 4) and also HMG1 is a well-known key enzyme in ergosterol biosynthesis. Furthermore, many recent studies comparing mRNA to protein levels in system biology consistently showed a correlation of only 0.4 to 0.6 (Geiger et al. 2013; Lundberg et al. 2010; Nagaraj et al. 2011). Nevertheless, the relatively poor correlation between mRNA and protein levels actually highlights the particular importance of proteome analysis in addition to more routine transcriptome analysis, and therefore we should not underestimate the importance of these proteins without good correlation versus their mRNAs.

Foc TR4 Proteomes for Early Development Provide an Important Clue for Screening New Fungicides

In the effort to combat Fusarium wilt of banana, one of the biggest challenges is the lack of effective fungicides in the field. Traditionally, a single fungicide used for controlling and preventing this disease in field has proven unsuccessful regardless of some effectiveness found for some individual fungicides in laboratory (Yang et al. 2006). This study provides the comprehensive proteomics analysis of Foc TR4 during early development. The results from characterization of the critical proteins in response to its conidial germination show an excellent consistence with the 6 target pathways being used for recently developed fungicides, as well as additional candidate targets, which are expected to shed light on and yield more choices for development of new antifungal drugs. Currently, the application of systemic fungicides, such as sterol demethylation inhibitors (DMIs), is essential for controlling Fusarium diseases to reach the attainable production level of modern high-yield cultivars. DMI-type fungicides, such as tebuconazole, propiconazole, triadimefon and prochloraz, act as ergosterol biosynthesis inhibitors to several key enzymes such as cytochrome P450 lanosterol C-14α-demethylase (ERG11/CYP51). Insufficient production of ergosterols is expected to subsequently disturb fungal membrane integrity (Yoshida 1993). In our study, several proteins in ergosterol synthesis pathways (ERG6, ERG11/CYP51, ERG13 and ERG25) appear to be the promising targets for further investigation and screening of new anti-fungal drugs. As a part of validation experiment, we have used five commonly-used fungicides including prochloraz as an ergosterol biosynthesis inhibitor and four other inhibitors irrelevant to ergosterol biosynthesis, for controls. Our results from in vitro growth inhibition of fungicides on the Foc TR4 demonstrated that prochloraz which suppresses ergosterol synthesis pathway, yielded more effective inhibition on Foc TR4 growth than the rest four did. This supports the findings in this proteomics profile study and provids another piece of evidence for the importance of ergosterol biosynthesis pathway in Foc TR4 early development. Thus, we believe that developing a new generation of fungicides or by dsRNAs with absorption of a selective action on ergosterol synthesis will become a new perspective for efficiently controlling Fusarium wilt of banana.

Taken together, the iTRAQ data, quantitative RT-PCR and western blot analyses have allowed us to identify important proteins and related pathways involved in Foc TR4 early development, including four enzymes, ERG6, EGR11, ERG13 and ERG25, in the ergosterol biosynthesis pathway. With follow-up in vitro growth inhibition analysis for validation of the identified key pathways, it may offer a promising new strategy for preventing and controlling Fusarium wilt. Overall, this work provides new insights into the Foc TR4 biological processes during its spore development, which could lead to the development of transgenic banana with efficient resistance to Fusarium wilt as well as new and more effective antifungal drugs against Foc TR4.