Telomere-to-telomere genome assembly of the aflatoxin biocontrol agent Aspergillus flavus isolate La3279 isolated from maize in Nigeria

ABSTRACT

Here, we report the complete genome of the non-aflatoxigenic Aspergillus flavus isolate La3279, which is an active ingredient of the aflatoxin biocontrol product Aflasafe. The chromosome-scale assembly clarifies the deletion pattern in the aflatoxin biosynthesis gene cluster and corrects a misidentified assembly previously published for this isolate.

ANNOUNCEMENT

Aspergillus flavus isolate La3279 (hereafter “La3279”) is a non-aflatoxigenic active ingredient of the aflatoxin biocontrol product Aflasafe (1). Previous work sequenced the La3279 genome and found 25 genes encoding aflatoxin biosynthesis (2). Another study analyzed the published La3279 genome and found that half of those genes were deleted (3). To resolve this inconsistency, we generated a new genome for La3279.

La3279 was grown from silica stock as described in reference (4). DNA was isolated and then sequenced on Oxford Nanopore Technologies and Illumina platforms following the methods detailed in reference (4). Quality filtering (QScore > 8) resulted in 2,097,937 nanopore reads (~7 gigabases). Read N₅₀ was 8.2 kilobases, calculated using BBTools v38.79 (sourceforge.net/projects/bbmap/). Nanopore reads were aligned to the final assembly using Minimap2 v2.24 (5), and the depth was estimated using SAMtools v1.9 (6). After library preparation and barcoding using NEBNext Ultra II DNA Library Prep Kit for Illumina, paired-end (150 nt) reads were sequenced on an Illumina NextSeq 550 at the Arizona Genetics Core, generating 27,771,766 reads (4.2 gigabases). Adaptor contamination was masked using Illumina software bcl2fastq. Quality assessment using FastQC v0.11.9 (7) revealed high-quality reads without adapter contamination. All read positions had median Phred scores > 26. Jellyfish v2.2.9 (8) generated a histogram of 21-mer counts, which was visualized using GenomeScope (9) (Fig. 1A). Canu v2.2 (10) assembly of nanopore reads generated 12 unitigs, which were polished with three rounds of Pilon v1.23 (11), realigning Illumina reads each time. One ~4 kilobase unitig was discarded because it matched bacteria. Two unitigs <4 kilobases and one unitig ~112 kilobases were discarded because they matched Aspergillus mitogenomes in BLASTN searches (12). The eight remaining unitigs, each longer than three megabases (Table 1), were numbered following homology with A. flavus NRRL3357 chromosomes (Fig. 1B; GenBank accession: GCA_014117465.1). The NCBI contamination screen was negative. An idiogram was generated using RIdeogram (13, 14). Synteny between La3279 and A. flavus NRRL3357 was visualized using D-GENIES (15). Chromosome ends were identified by manually searching for consecutive repeats of the dodecanucleotide “TTAGGGTCAACA” (16). GeneMark v4.59 was used with options “--ES” and “--fungus” to generate ab initio gene predictions (17). Funannotate v1.7.4 (18) predicted protein-encoding genes using options “--genemark_gtf” with GeneMark predictions and “--protein_evidence” with predicted proteins from A. flavus NRRL3357. The gene model for aflQ (ordA) required manual correction based on alignment with A. flavus NRRL3357. Annotation completeness was estimated using BUSCO (19). The La3279 assembly was aligned to the previously published La3279 assembly (GenBank accession GCA_008694415.1) using GSAlign v1.0.22 (20).

Fig 1.

Genome structure and comparison of A. flavus isolate La3279. (A) Blue bars under the curves represent the observed counts of k-mers of length 21 in Illumina sequencing reads. The orange line represents the modeled distribution of sequencing errors with low coverage. The black line represents the modeled distribution of k-mers across the La3279 genome. (B) D-GENIES whole-genome dot plot portraying synteny between A. flavus La3279 and A. flavus NRRL 3357. Dark green alignments correspond to the identity between 0.75 and 1, and light green near the end of chromosome 7 corresponds to the identity between 0.5 and 0.75. (C) Idiogram showing eight chromosomes of La3279 with a heatmap of gene density per 25 kb window.

TABLE 1.

Summary statistics of the Aspergillus flavus isolate La3279 genome assembly and annotations

	Aspergillus flavus isolate La3279
Size (Mb)	37.684
21-mer coverage	44.5×
Average depth of coverage, ONT long reads	~171×
Average depth of coverage, Illumina short reads	~107×
No. of chromosomes	8
N50 (Mb)	4.838
L 50	4
Largest scaffold (Mb)	6.56
GC content (% ±SD)	47.44 ± 0.2
Protein-encoding genes	12,727
BUSCO ascomycota	98.80%
BUSCO eurotiomycetes	98.80%
BUSCO eurotiales	98.40%
Chromosome 1 length (bp)	6,559,568
Chromosome 2 length (bp)	6,263,821
Chromosome 3 length (bp)	5,174,379
Chromosome 4 length (bp)	4,837,845
Chromosome 5 length (bp)	4,537,696
Chromosome 6 length (bp)	4,008,147
Chromosome 7 length (bp)	3,106,468
Chromosome 8 length (bp)	3,196,205

Telomeric repeats were found at 15/16 chromosome ends. Chromosomes of La3279 and A. flavus NRRL3357 showed 1:1 homology (Fig. 1B). Gene annotation predicted 12,727 protein-encoding genes (21) (Fig. 1C). The results of BUSCO analyses were in the 98th percentile (Table 1). At the end of chromosome 3, the aflatoxin biosynthesis gene cluster contained the 25 genes initially reported for La3279 (2). Our La3279 assembly shared 99.31% average nucleotide identity with the previously published assembly, suggesting that the original assembly belongs to a closely related isolate of A. flavus uploaded in error.

DATA AVAILABILITY

The genome and sequencing reads were deposited at NCBI under BioProject accession PRJNA992514. The BioSample accession is SAMN36357330. The SRA record for Illumina reads is SRX20939733. The SRA record for nanopore reads is SRX20939734. Genome and gene annotation files are available on figshare (20) (https://doi.org/10.6084/m9.figshare.23632116). This Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accession ASM3051527v1. The submitted GenBank assembly accession is GCA_030515275.1.