The killer toxin-like chitinases in filamentous fungi: Structure, regulation and potential functions in fungal-fungal interactions
Georgios Tzelepis1 and Magnus Karlsson2
1Department of Plant Biology, Uppsala BioCenter and Linnean Center for Plant Biology, Swedish University of Agricultural Sciences, Box 7080, S-75007 Uppsala, Sweden
2 Department of Forest Mycology and Plant Pathology, Uppsala BioCenter, Swedish University of Agricultural Sciences, Box 7026, S-75007 Uppsala, Sweden
* Corresponded author: Georgios Tzelepis: [email protected]
Keywords: chitinases, killer-toxins, LysM domains, mycoparasitism,
Plant and fungal cells are surrounded by a dynamic matrix, termed cell wall, which gives them the appropriate strength to protect themselves from variable environmental stress conditions, such as osmotic pressure, toxins, mechanical injures, while is also responsible for cell flexibility and rigidity (Bowman and Fee 2006, Keegstra 2010 Plant Physiol. 2010 Oct; 154(2): 483–486, Latge 2007 Mol Microbiol). In plant pathogens, molecules present in the cell wall are involved in host recognition and colonization (Huckelhoven 2007 Ann. Rev. Plant Pathol). Although the role of cell wall is similar between plant and fungal cells, their structure differs. Both are consisted of polysaccharides and glycoproteins, however, the main component in plant cell wall is cellulose (Keegsta et al 1973, Plant Physiology 51:188-197), while in fungi it is absent. In contrast, chitin is present especially in filamentous ascomycetes (Specht et al 1996 FGB). Chitin is a polymer consists of ?-1,4-linked N-acetyl-D-glucosamine (GlcNAc) and is the second most distributed polymer in nature after cellulose (Gooday 1990). It is the main component in crustacean and mollusc shells and in insect and nematode cuticle, while is also present in protozoa and algae (Mulisch 1993 Eur. J. Protist; Kapaun and Reisser Planta 197:577–582). Since chitin is absent from vertebrates and plants, makes it a perfect target for human drugs and pesticides (Spindler et al Parasitol. Res. 1990 76(4):283-8; Chaudhary et al 2013 Mini Rev. Med Chem. 13(2):222-36). In fungi, chitin is only the 10-20% in filamentous species, while in yeasts is even lower (1-2%) (Bartnicki-Garcia, S., 1968. de Nobel et al 2000 Klis et al 2002). Despite the fact that the percentage of chitin in fungal cell wall is relatively low its role is crucial for plasticity and rigidity (Specht et al 1996).
Chitin is degraded by chitinases (EC.126.96.36.199), which cleave the ?-1,4 glycan bond creating chitin oligomers or dimers (Gooday 1990). They are present in a wide range of organisms, eukaryotic and prokaryotic such as bacteria, algae, plant and fungi and play a crucial and vital role in chitin recycling in marine ecosystems (Keyhani and Roseman S 1999 Biochim. Biophys. Acta 1473:108–122). Fungal genomes contain a plethora of genes putatively encode chitin degrading enzymes (Seidl 2008). They play multiple roles in fungal life cycles, including cell wall remodelling and hydrolysis of exogenous chitin for nutrient acquisition (Baker et al 2009 Eukaryot. Cell, Dunker et al 2005 FGB, Boldo et al 2009 Curr Genet.), while mycoparasites utilize chitinases to attack to their fungal prey ( According to CAZy classification chitinases are grouped in glycoside hydrolases family 18 (GH18) and 19 (GH19) (Cantarel et al 2009). The latter are present in plant and bacteria species, while fungal chitinases have been only categorized in family 18 (Karlsson and Stenlid 2009). Furthermore, chitinases are clustered, regarding to their cleavage patterns, to exochitinases which cleave the chitin polymer from the exposed ends and to endochitinases which cleave the polymer randomly (van Alten et al 2000 PNAs, Horn et al 2006 FEBS J). Phylogenetically, GH18 fungal chitinases clustered in three distinct groups (A, B and C) which are further subdivided into several subgroups (Seidl et al 2005, Karlsson and Stenlid 2009).
Chitinases belong to group C subdivided in two groups CI and CII and display similarities with the ?/? subunit of the secreted toxin zymocin, produced by the dairy yeast Kluyveromyces lactis (Magliani et al 1997, Stark and Boyd 1986). These subunits display an exochitinases function and the exact role is to degrade the cell wall chitin from the prey species, facilitating the penetration of the ? subunit, which is the main toxin, to the cytoplasm (Butler et al 1991). The 3D model of C- group chitinases shows to be exochitinases as well (Gruber et al 2011 Glycobiology). The C group chitinases seem to be distributed in most of the filamentous fungi (Table 1 Number of C group chitinases in fungi), while an expanded number of genes, putatively encode for these chitinases, seem to be present in mycoparasitic Trichoderma species (Ihrmark et al 2010). Thus, the aim of this review article is to summarize the knowledge regarding the killer toxin-like chitinases in filamentous fungi and to propose a mode of action of these enzymes in fungal-fungal interactions.
2. 1 Domain structure
Most of the studied C group chitinases contain the DXXDXDXE motif in the GH18 catalytic domain, indicating that they are active chitinolytic enzymes (van Alten et al 2001, Gruber et al 2011, Tzelepis et al 2012, 2014). Furthermore, these proteins are putatively targeted to the endoplasmic reticulum, since a predicted signal peptide is present at the N-terminal (Gruber et al 2011, Tzelepis et al 2012, 2014). The most common feature of these chitinases is the presence of the CBM50 LySM domains. This domain shows a peptidoglycan binding affinity and first studied on prokaryotic organisms (Buist et al 2008, Mol Microbiol). Later it was described in many eukaryotic organisms such as plants and fungi (Zhang et al 2009 BMC Evol. Biol, Seidl et al 2005). In phytopathogenic fungi, LySM effector proteins play an important role in chitin binding, either protect fungal hyphae from plant chitinases, such as the Avr4 effector (van den Burg et al 2006 MPMI) or interfere in plant immunity, stealthing the hyphae from recognition such as the Ecp6 effector, in the hemi-biotrophic plant pathogen Cladosporium fulvum (de Jonge et al 2010 Science, van Esse et al 2006 MPMI). LySM domain proteins are widely spread not only in plant pathogens, but also in saprophytes (de Jonge and Thomma 2009 Trends Microbiol). In addition, the presence of a carbohydrate binding domain CBM18 (chitin-binding) has also been predicted in C-group chitinases.
Domain structure analysis on three Trichoderma species (T. atroviride, T. virens and T. ressei) (Gruber et al 2011 Glycobiology). In the first category, the GH18 catalytic domain is located almost in the middle of the protein, while one or two LySM and one CBM18 domains are located at the N-terminal (Figure 1). All T. ressei, which is a saprophytic species and shows contraction of killer-like toxin chitinase in its genome, contains four genes only in this group, while T. atroviride and T. virens contain two and eight respectively (Gruber et al 2011 Glycobiology). Regarding the model species Aspergillus nidulans and N. crassa, they contain four and two genes in that group respectively (Tzelepis et al 2012, 2014). The second category contains chitinase with only the CBM18 domain, and it is located together with the GH18 catalytic domain at the N-terminal of the protein (Figure 1). Phylogenetic analysis also showed that chitinases from the different modular categories are formed two distinct phylogenetic groups; C-I and C-II (Gruber et al 2011 Glycobiology). In addition, some chitinases in A. nidulans and N. crassa contain one or multiple transmembrane binding motifs at the C-terminal (Tzelepis et al 2012, 2014), indicating that these proteins are putatively localized to the plasma membrane (Figure 1).
Interestingly, Stergiopoulos et al (2012) identified that a small protein termed as Hce2, which is homolog to the C. fulvum Ecp2 effector, is fused to the C-terminal of some C group chitinases belong exclusively to the first modular class (Figure 1). This is effector has been identified in species such as Mycosphaerella graminicola and M. fijiensis (Stergiopoulos et al 2010 PNAs), and it seems to be a virulence factor (Lauge et al 1997 MPMI). Although the exact function of the Epc2 effector is not clear, it is speculated that it triggers necrosis in host plants (Stergiopoulos et al 2010 PNAs). It seems that this fusion associated more to non-pathogenic species, such as pathogens to humans, insects mycoparasites and saprophytes, rather than to plant pathogens (Stergiopoulos et al 2012). The role of the Hce2 protein is also unknown, but it could be a toxigenic peptide similar to the ?-subunit from the zymocin, playing a role in antagonist interactions.
2.3 Gene regulation
The transcription patterns of killer-toxin chitinases has been studied thoroughly in mycoparasitic and model species as well. In N. crassa, the gh18-6 and gh18-8 genes encoded for killer toxin-like chitinases with two LysM domains were highly induced during interactions with ascomycete plant pathogens Fusarium sporotrichioides as compared to carbon rich media, while the gh18-8 gene was also induced on carbon starvation conditions (Tzelepis et al 2012). However, different expression patterns were observed upon interactions with different fungal species. For instance, the gh18-6 gene was induced during interactions with the basidiomycete Rhizoctonia solani, while the other member of this subgroup the gh18-8 was induced during N. crassa self-interactions (Tzelepis et al 2012). Moreover, deletion of the gh18-6 gene has no impact in gh18-8 transcript levels (Tzelepis et al 2012). Regarding the third member of this group, different transcription patterns were observed. In particular, up-regulation was observed during carbon starvation conditions and not during fungal-fungal interactions (Tzelepis et al 2012).
Transcriptome analysis of C-II killer toxin-like chitinases in another filamentous model species Aspergillus nidulans has also been studied. The four members of C-II subgroup, all contain LySM and CBM-18 domains, were highly induced during interactions with B. cinerea and R. solani as compared to self-interactions, while no induction was observed upon interaction with Phytophthora which lacks chitin from their cell walls (Hardham 2007, Cell Microbial 9:31-39, Tzelepis et al 2014). Moreover, none of these genes were induced on chitin media, while down-regulation of them was observed on media where R. solani cell wall material was used. The transcriptome data derived from these two filamentous fungal model species, revealed that killer toxin-like chitinase potentially are involved in fungal-fungal interactions even in non-mycoparasitic species, while differential roles of these enzymes is also possible. Moreover, not only chitin, but also molecules produced by living prey hyphae are also required for induction of these chitinases.
The transcription patterns of C-group chitinase genes have been studied thoroughly in mycoparasites such as Trichoderma species. The first study was in T. atroviride by Seidl et al (2005), where induction of the chi18-10 gene, encode for a LysM contained chitinase, was observed before and after contact with R. solani mycelia and during growth on fungal cell wall material, in contrast to A. nidulans C-group chitinase genes (Tzelepis et al 2014). A more detailed analysis has been done by Gruber et al 2011. In this study
2.3 Potential mode of function