The genome of the butternut canker pathogen, Ophiognomonia clavigignenti-juglandacearum shows an elevated number of genes associated with secondary metabolism and protection from host resistance responses

DNA extraction and library preparation

For this study, the ex-type culture of Oc-j (ATCC 36624) isolated from an infected butternut tree in Wisconsin in 1978, was sequenced. For DNA extraction, isolates were grown on cellophane-covered potato dextrose agar for 7–10 d, and mycelia was collected and lyophilized. DNA was extracted from lyophilized mycelia using the CTAB method as outlined by the Joint Genome Institute for whole genome sequencing (Kohler A, Francis M. Genomic DNA Extraction, Accessed 12/12/2015 http://1000.fungalgenomes.org/home/wp-content/uploads/2013/02/genomicDNAProtocol-AK0511.pdf). The total DNA quantity and quality were measured using both Nanodrop and Qubit, and the sample was sent to the Hubbard Center for Genome Studies at the University of New Hampshire, Durham, New Hampshire. DNA libraries were prepared using the paired-end Illumina Truseq sample preparation kit, and were sequenced on an Illumina HiSeq 2500.

Genome assembly and annotation

We corrected our raw reads using BLESS 0.16 (Heo et al., 2014) with the following options: -kmerlength 23 -verify -notrim. Once our reads were corrected, we trimmed the reads at a phred score of 2 both at the leading and trailing ends of the reads using Trimmomatic 0.32 (Bolger, Lohse & Usadel, 2014). We used a sliding window of four bases that must average a phred score of 2 and the reads must maintain a minimum length of 25 bases. Next, de novo assembly was built using SPAdes-3.1.1 (Bankevich et al., 2012) with both paired and unpaired reads and the following settings: -t 8 -m 100 -only-assembler.

Genome sequences were deposited at ncbi.nlm.nih.gov under Bioproject number PRJNA434132. Gene annotation was performed using the MAKER2 pipeline (Holt & Yandell, 2011) in an iterative manner as is described in Yu et al. (2016), with protein evidence from related species of Melanconium sp., Cryphonectria parasitica, Diaporthe ampelina (from jgi.doe.gov) and Diaporthe helianthi (Baroncelli et al., 2016), for a total of three iterations. PFAM domains were identified in all the genomes using hmmscan with trusted cutoff (Bateman et al., 2004). Only nine species were included in the PFAM analysis. For the three Chrysoporthe species, D. aspalathi, and D. longicolla, protein sequences were not readily available for downloading online and e-mail requests were unsuccessful. Secondary metabolism gene clusters were identified using antiSMASH 4.0 (Weber et al., 2015).

Species phylogeny

Core eukaryotic proteins identified by CEGMA (Parra, Bradnam & Korf, 2007) were first aligned by MAFFT (Katoh et al., 2002) and then concatenated. Only proteins that were present in all genomes and all sequences were longer than 90% of the Saccharomyces cerevisiae ortholog were used. This resulted in a total of 340 single-copy genes that were concatenated into a single alignment. Phylogeny was then inferred using maximum likelihood by RAxML (Stamatakis, 2014) with PROTGAMMA used to estimate protein evolution, and with 100 bootstraps and midpoint rooted.

Secreted CAZyme prediction

SignalP (Petersen et al., 2011) was used to predict the presence of secretory signal peptides. CAZymes were predicted using CAZymes Analysis Toolkit (Park et al., 2010) based on the most recent CAZY database (www.cazy.org). Proteins that both contain a signal peptide and are predicted to be a CAZyme are annotated as a secreted CAZyme.

Genome assembly and annotation of Oc-j

The draft genome assembly of Oc-j contains a total of 52.6 Mb and 5,401 contigs, with an N50 of 151 Kbp (Table 1). The completeness of the genome assembly was assessed by identifying universal single-copy orthologs using BUSCO with lineage dataset for Sordariomycota (Simão et al., 2015). Out of 3,725 total BUSCO groups searched, we found that 3,378 (90.7%) were complete and 264 (7.1%) were fragmented in the Oc-j genome, while only 83 (2.2%) were missing. This result indicates that the Oc-j genome is relatively complete.

Genome-wide multi-gene phylogeny of Diaporthales

A total of 340 genes were used to generate a multigene phylogeny of the order Diaporthales. All sequenced species in Diaporthales (one genome per species for 12 species) were used. These species include: Cryphonectria parasitica (jgi.doe.gov), Chrysoporthe cubensis (Wingfield et al., 2015b), Chrysoporthe deuterocubensis (Wingfield et al., 2015b), Chrysoporthe austroafricana (Wingfield et al., 2015a), Valsa mali (Yin et al., 2015), Valsa pyri (Yin et al., 2015), Diaporthe ampelina (Savitha, Bhargavi & Praveen, 2016), Diaporthe aspalathi (Li et al., 2016), Diaporthe longicolla (Li et al., 2015), Diaporthe helianthi (Baroncelli et al., 2016), and Melanconium sp. The outgroups used were Neurospora crassa (Galagan et al., 2003) and Magnaporthe grisea (Dean et al., 2005). We show that the order Diaporthales is divided into two main branches. One branch includes Oc-j, C. parasitica, and the Chrysoporthe species (Fig. 2), in which Oc-j is located basally, indicating its early divergence from the rest of the branch. The other branch includes the Valsa and Diaporthe species.

Gene content comparison across the Diaporthales

To examine the functional capacity of the Oc-j gene repertoire, the PFAM domains were identified in the protein sequences (Table S1). For comparison, we also examined eight of the above mentioned 13 related species, where protein sequences were successfully retrieved (see “Methods” for details).

CAZymes are a group of proteins that are involved in degrading, modifying, or creating glycosidic bonds and contain predicted catalytic and carbohydrate-binding domains (Morales-Cruz et al., 2015). When secreted by fungal pathogens, CAZymes can participate in degrading plant cell walls during colonization; therefore, a combination of CAZyme and protein secretion prediction was used to identify and classify enzymes likely involved in cell wall degradation in plant pathogens (Morales-Cruz et al., 2015). Overall, the Oc-j genome contains 576 putatively secreted CAZymes, more than any of the other eight species included in this analysis, and 60 CAZymes more than D. helianthi, the species with the second most (Fig. 3; Table S1). Given that all except the bread mold species N. crassa, which has the lowest number of secreted CAZymes, are plant pathogens, this result indicates that Oc-j contains an especially large gene repertoire for cell wall degradation. However, it is not the greatest among members of the kingdom fungi. For instance, the saprophytic species Penicillium subrubescens and Podospora anserina have been found to have 719 and 590 CAZymes, respectively (Peng et al., 2017). This pattern also holds true when compared to other tree pathogens that have a latent period and an extended host range. For instance, Botryosphaeria dothidea, which is a pathogen of many woody species and has a prolonged latent and endophytic phase during the infection process, was reported to have 623 CAZymes (Marsberg et al., 2017). Other studies have also found an elevated number of CAZymes in multi-host pathogens of woody plant species including Eutypa lata and Neofusicoccum parvum which had 484 and 413 CAZymes, respectively (Morales-Cruz et al., 2015). Since secreted CAZymes are involved in degrading plant cell walls, the large number of secreted CAZymes in Oc-j and other pathogens of woody plants might help facilitate a life-style infecting multiple species of perennial woody plants with a wide array of both preformed and infection-induced resistance responses.

A total of 77 kinase related PFAM domains found in the nine above mentioned species. Of these, 61 domains are more abundant in Oc-j than the average of the eight related species, while only eight domains are less abundant in Oc-j (Fig. 3; Table S1). Given that kinases are involved in signaling networks, this result might indicate a more complicated signaling network in Oc-j. One example of kinase gene family expansion is the domain family of fructosamine kinase (PF03881). Oc-j contains 29 genes with this domain, while N. crassa has none and the other related species have between 3 and 8 genes (Fig. 3). Almost nothing is known about the function of fructosamine kinases in fungi. In other eukaryotes, it is involved in protein deglycation by mediating phosphorylation of fructoselysine residues on glycated proteins. While plants and fungi both produce an array of glycated proteins it is unclear if Oc-j is producing fructosamine kinase to function in deglycation of proteins produced by the plant host. Fructosamine kinases are extracellular serine-proteases, and are implicated in the pathogenic activity of several fungal pathogens of mammals (Da Silva et al., 2006), including Aspergillus fumigatus (Monod et al., 2002). Whether or not fructosamine kinase functions are important for plant pathogens remains unknown.

Cytochrome P450s (CYPs) are a superfamily of monooxygenases that play a wide range of roles in metabolism and adaptation to ecological niches in fungi (Chen et al., 2014). Due to their participation in a large number of detoxification reactions as well as in the metabolism of specific xenobiotics which may be co-assimilated as carbon source, CYPs are thought to be critical for the colonization of new ecological niches (Moktali et al., 2012). Among the nine species included in the PFAM analysis, N. crassa has only 31 CYPs, while the other eight plant-pathogenic species have between 104 and 223 CYPs. This result likely reflects that the ecological niche of N. crassa, saprphyte, is more favorable for fungal growth than alive plants with active defense mechanisms, thus fewer CYPs are needed for N. crassa to cope with relatively simple substrates. The Oc-j genome contains 223 CYPs (Fig. 3; Table S1), more than any of the other seven plant-pathogenic fungal species. This may be an evolutionary response to the number and diversity of secondary metabolites, such as juglone, produced by butternut as well as other Juglans species, that would need to be metabolized by any fungal pathogen trying to colonize the tree. Other pathogens of woody plants species are also known to have an increased number of CYPs. The grapevine canker pathogens E. lata and N. parvum have 205 and 212 CYPs respectively (Morales-Cruz et al., 2015). According to the Fungal Cytochrome P450 Database (FCPD) (Nelson, 2009), very few fungal species have more CYPs than Oc-j. The brown rot and pathogenic fungi in the basidiomycota have the greatest number of CYPs in the fungal kingdom. The brown rot fungus Postia placenta has the greatest number of CYPS with 345 CYPs, while the cacao pathogen Moniliophthora perniciosa and root rot pathogen Armillaria mellea have 300 and 245 CYPs, respectively (Nelson, 2009). The one characteristic all of these fungi have in common is their association with dead, dying or living trees and the need to breakdown or resist the multitude of chemical compounds produced by tree species.

To overcome host defenses, infect and maintain colonization of the host, fungi employ efflux pumps to counter intercellular toxin accumulation (Coleman & Mylonakis, 2009). Here, we examine the presence of two major efflux pump families, the ATP-binding cassette (ABC) transporters and transporters of the major facilitator superfamily (MFS) (Coleman & Mylonakis, 2009) in the Diaporthales genomes. ABC transporters (PF00005, pfam.xfam.org) are present in all nine species included in the PFAM analysis, ranging from 25 genes in N. crassa to 51 genes in V. mali, while Oc-j has 47 genes (Fig. 3; Table S1). The major facilitator superfamily are membrane proteins are expressed ubiquitously in all kingdoms of life for the import or export of target substrates. The MFS is a clan that contains 24 PFAM domain families (pfam.xfam.org). The Oc-j genome contains 653 MFS efflux pump genes, the most among all nine species included in the analysis (Fig. 3; Table S1). Cryphonectria parasitica has 552 genes, the second most. Interestingly, it was recently shown that the secondary metabolite juglone extracted from Juglans spp. could be used as potential efflux pump inhibitors in Staphylococcus aureus, inhibiting the export of antibiotics out of the bacterial cells (Zmantar et al., 2016). In line with this previous finding, our results suggest that Oc-j genome contains a large arsenal of efflux pumps likely due to the need to cope with hostile secondary metabolites such as the efflux pump inhibitor juglone.

Secondary metabolism gene clusters. Secondary metabolite toxins play an important role in fungal nutrition and virulence (Howlett, 2006). To identify gene clusters involved in the biosynthesis of secondary metabolites in Oc-j and related species, we scanned their genomes for such gene clusters. We found that the Oc-j genome contains a remarkably large repertoire of secondary metabolism gene clusters, when compared to the closely related C. parasitica and the Chrysoporthe species (Fig. 3; Table S1). While C. parasitica has 44 gene clusters and the Chrysoporthe species have 18–48 gene clusters, the Oc-j genome has a total of 72 gene clusters (Table S1), reflecting a greater capacity to produce various secondary metabolites. Among the 72 gene clusters in Oc-j, more than half are type 1 polyketide synthases (t1pks, 39 total, including hybrids), followed by non-ribosomal peptide synthetases (nrps, 14 total, including hybrids) and terpene synthases (9) (Table S1).

When compared to all species included in this study, we found that Oc-j has the third highest number of clusters, only after D. longicolla and Melanconium sp.; however, when normalized by total gene number in each species, Oc-j is shown to have 6.4 secondary metabolism gene clusters per 1,000 genes, the highest among all species included in this study. The D. helianthi genome contains only six clusters, likely due to the poor quality of the genome assembly with an N50 of ~6 kb (Fig. 3; Table S1). Cluster numbers can vary greatly within the same genus.