Genome Resource of Colletotrichum spaethianum, the Causal Agent of Leaf Anthracnose in Polygonatum falcatum

Colletotrichum spaethianum is a fungal phytopathogen causing leaf anthracnose that is phylogenetically related to the root endophyte C. tofieldiae within the spaethianum species complex of Colletotrichum spp. fungi. However, limited studies of this fungus have rendered its pathogenesis and host interaction elusive. Here, for the first time, we characterized draft genome sequences of C. spaethianum MAFF 239500, which causes leaf anthracnose in Polygonatum falcatum A. Gray in Japan. This study will provide insight into the genomic potential and serve as a material to get a better understanding of the diverging traits between related phytopathogenic and endophytic fungi.

The ascomycetes from the genus Colletotrichum have been noted as phytopathogens that cause anthracnose diseases in various economically important plants worldwide (Dean et al. 2012). C. spaethianum has been identified as a cause of this disease on plants from genus Polygonatum, such as P. falcatum (also called narukoyuri, a popular market flower in Japan), P. odoratum (an ornamental flowering plant), and P. cyrtonema (a traditional Chinese herb) (Liu et al. 2020; Ma et al. 2021; Tomioka et al. 2008). For example, C. spaethianum MAFF 239500 (originally identified as C. dematium) causes chlorotic to brown spots on P. falcatum leaves at the initial infection stage, followed by severe lesions, making the leaves finally blighted and defoliated (Sato et al. 2015; Tomioka et al. 2008). Interestingly, C. spaethianum has a close phylogenetic relationship to C. tofieldiae, a beneficial endophyte that promotes plant growth of the model plant Arabidopsis thaliana plants by transferring phosphorus to the host under phosphate-limiting conditions (Hiruma et al. 2016). These fungal species, along with other phytopathogenic species such as C. incanum and C. liriopes, form the spaethianum species complex within the genus Colletotrichum phylogenetic tree (Talhinhas and Baroncelli 2021). However, the molecular bases underlying the lifestyle differences in those closely related species are hitherto unknown. As the first step to elucidate these mechanisms, we present the first draft genome and annotation of C. spaethianum, which causes diseases on P. falcatum.

Fungal colonies of C. spaethianum MAFF 239500 were isolated from diseased leaves of narukoyuri (P. falcatum A. Gray) grown in open fields in Kagawa Prefecture, Japan (Tomioka et al. 2008). Mycelia for DNA extraction was obtained by culturing agar fragments containing fungal colonies in liquid Mathur's medium incubated on a rotary shaker for 100 rpm at 25°C for 3 days. The genomic DNA was subsequently extracted by the cetyltrimethylammonium bromide method as described by Damm et al. (2008). Equimolar of DNA samples were sent for long-read sequencing on a PacBio RSII system in Macrogen Corp., Japan. DNA samples were also used to identify ITS, ACT, CHS-1, TUB2, HIS3, and GAPDH genes by PCR for phylogenetic assessment, as described by Cannon et al. (2012). The phylogenetic position of MAFF 239500 as C. spaethianum within the spaethianum species complex was confirmed (Supplementary Fig. S1) by maximum-parsimony analysis using PAUP v.1.3.3 with a heuristic search option (Swofford 2003).

Genome assembly and annotation statistics are summarized in Table 1. Briefly, pre-assembly filtering and de novo assembly for generated reads were performed using Falcon v.2.1.4 (Chin et al. 2016), resulting in a total sequence of 50.9 MBp in 84 scaffolds (Table 1). Draft-genome completeness was measured using BUSCO v.5.1.2 (Simão et al. 2015) based on the glomerellales_odb10 lineage dataset, resulting in 92.6% genome completeness. Repetitive regions were detected as 1.2% of the genome using Dfam TETools containing RepeatMasker v.4.1.2-p1 with fungi as species option (Storer et al. 2021). Sequence identity to phylogenetically close species was explored using Mauve whole-genome alignment using the MCM algorithm (Darling et al. 2004). The longest locally colinear block formed between C. spaethianum and C. tofieldiae (0861 strain, GenBank accession: GCA_001625265.1) was 1.26 Mb with 71.2% identity, with C. liriopes (A2 strain, GenBank accession: GCA_015832465.1) it was 0.92 Mb with 72% identity, and with C. incanum (MAFF238704, GenBank accession: GCA_001625285.1) it was 0.72 Mb with 66.4% identity.

Table 1. Summary of genome sequencing, assembly, and annotation statistics

Feature	Colletotrichum spaethianum MAFF 239500
Raw data (PacBio RS2)	4,870,594,510
Assembly software	FALCON
Genome (bp)	50,915,707
Total sequence coverage	63×
Number of scaffolds	84
Largest scaffold (bp)	3,303,556
Smallest scaffold (bp)	3,502
Genome N50 (bp)	1,120,470
GC content (%)	51.1
BUSCO completeness (genome mode) (%)	92.6
BUSCO completeness (protein mode) (%)	85.3
Repeat rate (%)	1.2
Number of genes	13,262
Number of coding sequences (CDS)	12,842
Average of CDS length (aa)	416
Average of transcript length (bp)	1,250
Average of exons per gene	2
Number of exon	30,966
Average of exon length (bp)	518
Number of intron	18,124
Average of intron length (bp)	84
Number of transfer RNA	355
Number of ribosomal RNA	65
Secreted proteins	1,245
Candidate secreted effector proteins	428
Carbohydrate-active enzyme	629
Candidate secondary metabolites clusters	42

View as image HTML

Protein and coding sequence (CDS) data of C. tofieldiae (Hacquard et al. 2016) were used as training data for SNAP v2.31.8 (Korf 2004). The resulting gene training model from SNAP was used for gene prediction using MAKER v.2.31.8 (Cantarel et al. 2008), resulting in 12,842 CDS. Subsequent functional annotation was performed using Protein BLAST+ v.2.6.0 (Camacho et al. 2009) based on the UniProt Swiss-Prot database (June 2018 version). Detection and annotation of transfer RNA (tRNA) and ribosomal RNA (rRNA) were conducted using tRNA-scan v.2.0.7 (Lowe and Chan 2016) and barrnap v.0.9 (Seeman 2018), resulting in 355 tRNAs and 65 rRNAs, respectively. Prediction of secretome was conducted as described by Crestana et al. (2021) using SignalP v.5.0 (Almagro Armenteros et al. 2019), TMHMM v.2.0 (Krogh et al. 2001), and ProSite PS-Scan against motif PS00014 (de Castro et al. 2006). From 1,245 candidate secretomes, 428 candidate effector proteins were detected using EffectorP v.3.0 (Sperschneider and Dodds 2022). Additionally, identification of 629 carbohydrate-active enzymes (CAZymes) was performed with the hmmscan function from HMMER v.3.1b2 (Eddy 2011) using dbCAN HMMdb release 9.0 (Zhang et al. 2018) as a database. From these CAZymes, 127 genes are categorized under the ligninolytic and lytic polysaccharide mono-oxygenases family, 15 genes under the noncatalytic carbohydrate-binding modules family, 54 genes under the carbohydrate esterases family, 308 genes under the glycoside hydrolases family, 91 genes under the glycosyltransferases family, and 34 genes under the polysaccharide lyases family. We also detected 42 gene clusters as candidates of secondary metabolites using anti-SMASH v.6.0.1 (Blin et al. 2021). The numbers of these effectors, CAZymes, and secondary metabolite clusters are comparable with previously sequenced species in the same spaethianum species complex with higher genome completeness (Supplementary Table S1).

To the best of our knowledge, this is the first draft genome sequence of C. spaethianum, which is a pathogenic fungus causing leaf anthracnose in P. falcatum. We expect this draft genome will provide valuable information for future comparative studies of the spaethianum species complex, including pathogenic and endophytic Colletotrichum spp., and broaden our understanding of plant and fungal molecular interaction.

Data and Material Availability

The assembled contigs and annotations generated in this study were deposited in the DNA Data Bank of Japan under the BioProject PRJDB11870 and BioSample SAMD00334500 with contig accession numbers BQXU01000001 to BQXU01000084. Fungal isolates material is available at NARO Genebank, Japan as isolate MAFF 239500.

Note Added in Final Publication

Changes were made to this article just prior to final publication to reflect new information. Previous information for C. spaethianum reported here had been switched with that of a closely related Colletotrichum species (C. liriopes). The whole-genome sequencing was carried out at the same time with the same procedure.