Genome Diversity, Population Structure and MALDI-TOF MS Profiling of Aspergillus oryzae/flavus Strains from Fermentation and Wild Environments

doi:10.21203/rs.3.rs-4865798/v1

Download PDF

Research Article

Genome Diversity, Population Structure and MALDI-TOF MS Profiling of Aspergillus oryzae/flavus Strains from Fermentation and Wild Environments

https://doi.org/10.21203/rs.3.rs-4865798/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Various strains of Aspergillus oryzae, regarded as a domesticated variant of aflatoxigenic Aspergillus flavus, are utilized in soybean fermentation industry of Korea. This study compared A. oryzae/flavus strains isolated from various environments in Korea, including industrial settings, Meju (brick of dried fermented soybeans), and wild conditions, with globally reported strains using genomic analysis to determine their taxonomic positions and safety. Using population genomics, five distinct groups (A to E) were identified, with all aflatoxigenic Korean strains in Group C and non-aflatoxigenic Korean strains in Groups A, B, and E. Korean strains from Meju and wild conditions are distributed across Groups A and B, and most of the Korean industrial strains form a sub-cluster with Japanese industrial strains in Group A. Comparing secondary metabolite gene cluster mutation pattern, three gene clusters (Aflatoxin, Cyclopiazonic acid and Ditryptophenaline) were revealed as group specific ones. In aflatoxin and cyclopiazonic acid clusters, most of the Group C strains had intact regions compared to other groups strains.Since most of the Group C strains produce aflatoxin and have intact Aflatoxin and Cyclopiazonic acid gene cluster, we considered that this group represent A. flavus. Profiling of MALDI-TOF MS analysis also distinguished Group C from Groups A, B and E by specific proteomic peaks. Among these peaks, those around 12700 to 12900 m/z (Da) are expected to correspond to AflF (nor B), an enzyme involved in Aflatoxin metabolism. These results showed taxonomic positions of Korean strains of A. oryzae/ flavus from various environments and also showed possibility to differentiate between A. oryzae and A. flavus with genome and Maldi-tof analysis.

genomics

Aspergillus oryzae/flavus complex

structure analysis

aflatoxin

MALDI-TOF

In Asia, the non-aflatoxigenic filamentous fungal species Aspergillus oryzae plays an indispensable role in the fermentation process of various foods and beverages, including doenjang, soy sauce, and sake [1–5]. This species is considered a domesticated variant of Aspergillus flavus, a species notorious for producing harmful mycotoxins such as aflatoxin and cyclopiazonic acid. [3, 6].

A previous study demonstrated that the A. oryzae strain RIB40 is incapable of producing secondary metabolites, aflatoxin, cyclopiazonic acid and produces isomeric metabolites such as aflatrem, miyakamides and ditryptophenaline compared to A. flavus strain NRRL 3357 [7]. Despite their significant biochemical differences, these two species exhibit a close genomic relationship. Moreover, certain strains of A. flavus exhibited variability in aflatoxin production, influenced by temperature and humidity [8]. Due to these characteristics, some strains were classified as A. flavus, despite not producing aflatoxin [9]. Therefore, comparing whole genomes is required to differentiate these closely related species precisely.

A previous whole genomic analysis of A. oryzae classified industrial strains used in the Japanese fermentation industry into eight clades [5]. Further efforts aimed at distinguishing A. oryzae from A. flavus through comparative genomics, with a particular focus on aflatoxin gene cluster and the composition of Carbohydrate-Active enzymes (CAZymes), have not conclusively differentiated the two species [10].

While extensive genomic studies have been conducted on industrial strains, relatively little attention has been given to non-aflatoxigenic strains derived from traditional meju and wild conditions, which are expected to have fewer events of genetic improvement. In a study on Penicillium roqueforti, traditional strains used in Roquefort cheese production were found to be intermediate between wild and industrial strains. Thus, strains are also presumed to be intermediate between industrial A. oryzae and aflatoxigenic A. flavus [11]. Therefore, understanding their population structure and aflatoxin production risk at the genomic level is necessary to monitor their distribution.

In recent microbial identification and classification study, the MALDI-TOF MS (Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry) method has emerged as a powerful tool. For example, Weissella confusa and Weissella cibaria are genetically close and had different roles, with W. confusa contributing to health benefits while W. cibaria having a potential pathogenic role. This relation is similar to relationship between A. oryzae and A. flavus. To distinguish these closely related species, recent studies have developed markers using machine learning to analyze MALDI-TOF MS patterns [12]. There was also attempt to use MALDI-TOF MS to distinguish A. oryzae/flavus complex, but the study concluded that this method is not suitable for accurate differentiation due to the high overlapping protein profiles of the two species [13].

Our study aimed to elucidate the distribution of diverse Korean A. oryzae/flavus strains from various environments by comparing their genomes with those of strains from previous studies originated from diverse geographical regions, such as Japan, the United States, and China.

Furthermore, variations of mutations within gene clusters associated with mycotoxins and secondary metabolites, related to safety in the food industry, were analyzed. In addition, MALDI-TOF MS for proteomics pattern profiling was conducted to distinguish between clustered groups. Based on the findings, we propose group distribution of A. oryzae and A. flavus, and their specific characteristics.

2.1. Collection of Genomes of Korean and Global A. oryzae/flavus Strains

Korean strains from 4 diverse sources, Korean Industrial non-aflatoxigenic strains (KRI, 5 strains), Korean non-aflatoxigenic strains from meju (KRM, 12 strains), Korean non-aflatoxigenic strains from wild conditions (KRWO, 7 strains), Korean aflatoxigenic strains from wild conditions (KRWF, 15 strains) were collected (Table 1). All 39 Korean strains were stored at -80°C. A portion of each stock was cultured on Malt Extract Agar (MEA) at 25°C for seven days for genomic DNA extraction.

Table 1

Korean A. oryzae/flavus strains from various condition. KRM are non-aflatoxigenic strains from traditional meju.KRI are non- aflatoxigenic strains used in Korean fermentation Industry. KRWO are non- aflatoxigenic strains from Korean wild condition. KRWF are aflatoxigenic strains from Korean wild condition.
Strain name		KACC no.	Discription	Region	Aflatoxin	norB/cypA	omtA
MWA1	KRM	K46457	Meju, Goisan, omtA A type (Hong et al. (2013)	Korea	X	Type I	IA-S-B+
MWA2		K46810	Meju, Sunchang, omtA A type (Hong et al. (2013)	Korea	X	X	IA-L-B-
MWA3		K46811	Meju, Jeju, omtA A type (Hong et al. (2013)	Korea	X	X	IA-L-B-
MWB1		K46455	Meju, Gyeongsan, omtA B type (Hong et al. (2013)	Korea	X	Type I	IB-L-B-
MWB2		K46470	Meju, Yongin, omtA B type (Hong et al. (2013)	Korea	X	Type I	IB-L-B-
MWB3		K46471	Meju, Incheon, omtA B type (Hong et al. (2013)	Korea	X	X	IB-L-B-
MWC1		K46456	Meju, Haenam, omtA C type (Hong et al. (2013)	Korea	X	Type I	IC-Ao
MWC2		K46474	Meju, Damyang omtA C type (Hong et al. (2013)	Korea	X	Type I	IC-Ao
MWC3		K46469	Meju, Jeju, omtA C type (Hong et al. (2013)	Korea	X	Type I	IC-Ao
MWX1		K46465	Meju, Buan, no omtA(Hong et al. (2013)	Korea	X	Type I	X
MWX2		K46466	Meju, Gongju, no omtA(Hong et al. (2013)	Korea	X	Type I	X
K93210		K93210	Meju, patent strain	Korea	X	Type I	IC-Ao
KRI1	KRI	K46640	Korean industrial strain	Korea	X	Type I	IC-Ao
KRI2		K46641	Korean industrial strain	Korea	X	Type I	IB-L-B-
KRI3		K46642	Korean industrial strain	Korea	X	Type I	IC-Ao
KRI4		K47488	Korean industrial strain	Korea	X	Type I	IC-Ao
KRI5		K47843	Korean industrial strain	Korea	X	Type I	IC-Ao
Aor-06	KRWO	K46923	Rice straw, Yangyang	Korea	X	Type I	ND
Aor-17		K46924	Rice straw, Gongju	Korea	X	Type I	ND
Aor-38		K46909	Soybean, Incheon	Korea	X	X	ND
SD045		K48145	Peanut, Jinyang	Korea	X	X	ND
SL041		K46918	Soil, Suwon	Korea	X	Type I	ND
SL046		K46920	Soybean farm soil, Namhae	Korea	X	Type I	ND
SL055		K46922	Greentea farm soil, Jeju	Korea	X	X	ND
Aor-34	KRWF	K46927	Rice straw, Yangpyeong	Korea	O, B	Type II	IC-L-B+/B-
AR018		K46892	Indoor air, Seoul	Korea	O, B	Type II	IC-L-B+/B-
AR028		K46894	Outdoor air, Suwon	Korea	O, B/G	Type III	Out group
SD016		K46897	Corn, Hongcheon	Korea	O, B/G	Type III	Out group
SD022		K46898	Corn, Hongcheon	Korea	O, B	Type II	IC-L-B+/B-
SD024		K46899	Corn, Hongcheon	Korea	O, B	Type II	IC-L-B+/B-
SD039		K46902	Peanut, Seosan	Korea	O, B/G	Type II	IC-L-B+
SD059		K46903	Peanut, Jangseong	Korea	O, B	Type II	IC-L-B+/B-
SD061		K46904	Peanut, Danyang	Korea	O, B/G	Type II	IC-L-B+/B-
SL001		K46928	Soil, Gwangju	Korea	O, B	Type II	IC-L-B+
SL005		K46929	Soil, Gwangju	Korea	O, B	Type II	IC-L-B+
SL008		K46912	Soil, Deagu	Korea	O, B	Type II	IC-L-B+/B-
SL015		K46913	Soil, Gwangju	Korea	O, B/G	Type III	Out group
SL034		K46916	Soil, Suwon	Korea	O, B/G	Type II	IC-L-B+/B-
SL044		K46919	Garlic farm soil, Namhae	Korea	O, B/G	Type II	IC-L-B+/B-

The genomic DNA (gDNA) from Korean strains was extracted using DNeasy Plant Mini kit (Qiagen, Hilden, Germany) and its protocol. PCR-free 150-bp paired-end libraries were constructed and sequenced by Macrogen Inc (Seoul) on an Illumina NovaSeq 6000. Raw reads underwent quality control. Trimmomatic (v0.38) was used to remove adapter sequences and low-quality reads [14] and FastQC was performed to check the read quality (https://www.bioinformatics.babraham.ac.uk/projects/fastqc).

For the comparative genomic analysis, sequence reads from strains originated from 4 different sources were collected. The sources include Japanese industrial strains (JPI, 17 strains), NCBI non-aflatoxigenic strains (NO, 14 strains), NCBI aflatoxigenic strains (NF, 8 strains), and Afla-guard non-aflatoxigenic strains (AG, 2 strains). All sequence reads were obtained as raw files from the National Center for Biotechnology Information (NCBI) Sequence Reads Archive (Table 2).

Table 2

Global A. oryzae/flavus strains from NCBI database. JPI are atoxigenic strains used in Japanese industry. Two strains for each clades, 8 clades strains are gathered. TK-24 is not included in any clade. NO are global strains known as atoxigenic. NF are global strains known as toxigenic. AG are strains known as Afla-Guard.
Strain name		NCBI No.	Discription	Region	Aflatoxin
TK-1	JPI	DRX154124	Sake, Clade A in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-2		DRX154126	Sake, Clade A in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-20		DRX154161	Sake, Clade B in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-26		DRX154172	Sake, Clade B in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-24		DRX154168	Outer Clade in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-5		DRX154132	Sake, Clade C in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-7		DRX154136	Miso, Clade C in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-59		DRX154211	Soyou, Clade D in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-60		DRX154212	Soyou, Clade D in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-9		DRX154140	Soyou, Clade E in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-13		DRX154148	Soyou, Clade E in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-4		DRX154130	Sake, Clade F in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-11		DRX154144	Miso, Clade F in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-10		DRX154142	Miso, Clade G in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-15		DRX154151	Miso, Clade G in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-12		DRX154146	Soyou, Clade H in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
TK-14		DRX154149	Soyou, Clade H in Naoki et al. (2019)	Japan	X, Naoki et al. (2019)
14160	NO	SRX013842	Xinyang City, Chacón-Vargas et al. (2021),	China	X, Chacón-Vargas et al. (2021)
BP2-1		SRX6074494	Korean strain in Naoki et al. (2019)	Korea	X, Naoki et al. (2019)
RIB537		SRX147127	Gibbons et al. (2012)	America	X, Gibbons et al. (2012)
RIB949		SRX147131	Gibbons et al. (2012)	America	X, Gibbons et al. (2012)
M2040 *		SRX4479827	Biocontroller in Alshannaq et al. (2018)	Korea	X, Alshannaq et al. (2018)
NRRL35739		SRX5329434	Biocontroller in Pennerman et-al. (2019)	America	X, Pennerman et-al. (2019)
SU-16		SRX8635871	Suzhou winery	China	X, Sun, Liu et al. (2022)
WRRL1519		SRX3067799	almond nuts	America	X, Yin et-al. (2018)
2017 Washington T2		SRX5358295	Biocontrol strain	America	X, Fountain et al. (2020)
A1		SRX8062526	Louisiana State University	America	X, Fountain et al. (2020)
AF36		SRX8062528	Biocontrol strain	America	X, Fountain et al. (2020)
K54A		SRX8062531	Louisiana State University	America	X, Fountain et al. (2020)
NRRL30797 (K49)		SRX8062530	Biocontrol strain	America	X, Fountain et al. (2020)
VCG1		SRX8062533	Louisiana State University	America	X, Fountain et al. (2020)
2017 Washington T5	NF	SRX5358298	USDA-ARS	America	O,
A9		SRX8062527	Louisiana State University	America	O, Fountain et al. (2020)
CA14		SRX6432310	Pistachio, Hua et al.(2012}	America	O, Hua et al. (2012)
E1445		SRX8621731	Ethiopia peanut	Ethiopia	O, Arias,Mohammed et al. (2020)
E1404		SRX8617991	Ethiopia peanut	Ethiopia	O, Arias,Mohammed et al. (2020)
E1402		SRX8617924	Ethiopia peanut	Ethiopia	O, Arias,Mohammed et al. (2020)
MRI19		SRX10945235	Tiger nuts, Schamann et al. (2022)	Spain	O, Schamann et al. (2022)
Tox4		SRX8062532	Louisiana State University	America	O, Fountain et al. (2020)
Afla-Guard	AG	SRX4479828	University of Georgia	America	X, Fountain et al. (2020)
Yazoo S2	AG	SRX5356519	USDA-ARS(Afla guard)	America	X, Fountain et al. (2020)
* : Non-aflatoxigenic strains from traditional meju; KRM strain.

JPI strains, previously categorized into eight clades (Clade A to H), were chosen with two representatives from each clades and an extra strain that did not fit into any clade (TK-24) were selected [5].

The distinction between the NO and NF strains was not based on the existing nomenclature. Instead, it was determined by previous research studies that examined the aflatoxin production capabilities of these strains [3–5, 9, 15–21].

2.2. Single Nucleotide Polymorphisms (SNPs) calling and population genetic analysis

For SNP calling, filtered DNA sequencing reads of the 80 individuals were aligned to the reference genome, NRRL 3357, using BWA software v0.7.17 MEM module [22]. SAM alignment files were sorted and converted to BAM files with SAMtools v1.3 [23]. Picard was used to delete duplicat reads and create index of reference genome. Next, various programs in GATK packages (HaplotypeCaller for variant calling, CombineGVCFs for merging GVCF files, GenotypeGVCFs for converting GVCF into VCF file, SelectVariants, and VariantFiltration for selecting and filtering SNPs) were used [24]. The filtering parameters were like this: “QD < 2.0 | | FS > 80.0 | | MQ < 20.0| | SOR > 3.0| | MQRankSum < − 12.5| | Read- PosRankSum < − 8.0 | | QUAL < 40.0”. VCF files were converted into fasta files using vcf2fasta to construct a phylogenetic tree (https://github.com/santiagosnchez/vcf2fasta). The raw data of Illumina squences were deposited on NCBI Sequence Reads Archive.

The tree was drawn using whole-genome-based phylogenetic reconstruction program, SANS serif [25]. PCA with population-scale SNPs was performed using plink [26] and GCTA [27]. A Bayesian population structure assessment was performed using ADMIXTURE v1.3.0 [28] with block relaxation algorithm, maximum likelihood estimation. Pre-defined genetic clusters were increased from K = 2 to K = 10 (assumed number of ancestral populations). Prior to running ADMIXTURE, SNPs in high linkage disequilibrium (LD) were removed. Specifically, PLINK was used to prune SNPs employing a windowed approach, with an R2 threshold of 0.8.

2.3. Genotypic Analysis of Secondary Metabolites Gene Cluster and Design of Diagnostic Markers

Significant phenotypic differences influenced by A. oryzae/flavus domestication are prominently reflected in the mutations of secondary metabolites, with substantial variations in the secretion levels of numerous toxic metabolites. The presence and locations of the aflatoxin gene cluster [29–31], the CPA gene cluster [32], and the ditryptophenaline gene cluster [33] in the genome of Aspergillus flavus NRRL 3357 have been elucidated in previous studies. SNPs based on the NRRL 3357 genomic sequence are visualized using the Integrated Genomics Viewer (IGV). This visualization allows the identification of specific mutations or deletions occurring within these gene clusters of each strain.

2.4. MALDI TOF metabolites pattern analysis

Mass spectrometry data from MALDI-TOF/MS on KACC strains containing Groups A, B, C and E were analyzed to compare the similarity patterns among groups classified based on genomic analysis results. The group specific mass-spectral loci were explored by comparing presence or absence of mass peak at certain loci and their degree of difference.

Each strain underwent individual mass analysis using MicroIDSys® (ASTA Inc.). A sample from each strain is processed by mixing it with a cell lysis solution. After cell lysis, 1.5 µL of the separated supernatant is applied to the µID plate, dried, and then coated with 1.5 µL of CHCA matrix. The raw spectrum data generated by MicroIDSys® ranges from 2,000 to 20,000 daltons (m/z), with an identification cutoff set at ≥ 140 as per the manufacturer’s recommendation.

The raw data from MicroIDSys® undergoes normalization to adjust peak height relative to m/z using proprietary software from ASTA Inc. and NQ-Lab. Co.,Ltd. Subsequently, the data is processed with a binning parameter set at 5. An averaging procedure calculates the mean peak heights for each m/z range within the binning. The processed data is then interpreted using a Python-based code, resulting in a combined dataset.

The prepared data is analyzed using a machine learning model developed by NQ-Lab., Co., Ltd. Logistic Regression, is the chosen machine learning algorithm tailored with hyperparameters optimized for MALDI-TOF/MS characteristics. The analysis accounts for various factors influencing MALDI-TOF/MS results, such as temporal, physical, and biochemical attributes. To ensure a comprehensive analysis, multiple statistical tests are conducted. The ANOVA test is utilized to examine mean differences between groups, considering the continuity of m/z values. Additionally, the Chi-squared (χ²) test is employed to assess the independence of categories, assuming that the two groups are distinct entities.

3.1. Phylogenetic relationship of A. oryzae/flavus strains

This comprehensive analysis classified the non-aflatoxigenic and aflatoxigenic A. oryzae/flavus strains into five distinct groups, designated as Groups A to E (Fig. 1).

Group A predominantly comprises non-aflatoxigenic strains, with a significant presence of KRI and JPI strains. Previous research by Watari et al. (2019) categorized JPI strains into eight clades (Clade A to H), and strains belonging to Clades C to H were integrated into Group A. Notably, most KRI strains were found to cluster within Clade C. Additionally, among the KRM strains, MWC2, MWC3, and M2040 also aligned with Clade C, while K93210 aligned with Clade D. Other KRM and KRWO strains were also classified under Group A, although they did not align with any of the specific clades. Group B primarily includes non-industrial strains such as KRM, KRWO, and NO, apart from a singular KRI strain, KRI2, indicating a diverse genetic background within non-aflatoxigenic populations.

Group C encapsulates all aflatoxigenic strains, including KRWF and NF. Intriguingly, this group also contains six non-aflatoxigenic strains from the NO category (2017 Washington-T2, A1, AF36, K54A, NRRL30797, and VCG1), highlighting the complex genetic landscape that does not strictly correlate with aflatoxin production.

Group D is characterized by its genetic divergence from Groups A, B, and C, comprising mainly JPI strains that fall into the A and B clades as defined by Watari et al., with the exception of one NO strain, 14160.

Group E, distinguished as the most genetically distinct cluster, includes biocontrol strains such as Afla-guard and Yazoo S2, alongside two unique KRM strains (MWA1 and MWA3).

3.2. Genome and Population Structure Analysis of A. oryzae/flavus

Principal Component Analysis (PCA) of the A. oryzae/flavus population structure revealed that Group D closely aligns with Groups A and B, demonstrating their closer genetic relationship, while Group E exhibits distinct genetic separateness from all other groups (Fig. 2A). This distinct positioning of Group E highlights its unique genetic makeup.

Admixture analysis further elucidates the genetic diversity and interrelation among the groups. To identify the most optimal K number, we calculated the cross-validation (CV) error values, where a lower CV value indicates a better estimation. This calculation was performed systematically for K values ranging from 1 to 15 [34]. Using this approach, K = 11 was identified as the most likely number of populations. At K = 11, part of the Group A strains shared population with Group B and D, while part of the Group C strains shared small population with Group E. Phylogenetic network analysis further supports the presence of five main populations and individual population assignment into these populations (Fig. 2B).

3.3. Secondary Metabolite Gene Cluster Analysis

Among the various secondary metabolite gene clusters, three metabolites gene clusters, aflatoxin, CPA and ditryptophenaline gene clusters showed group-specific characteristics. In genome of A. flavus NRRL3357, aflatoxin gene cluster region is located on chromosome 3, 4,940,000 to 5,010,000 (Fig. 3). CPA gene cluster region is located on beside the aflatoxin gene cluster, chromosome 3, 5,010,000 to 5,034,262 (Fig. 4). The ditryptophenaline gene cluster region is located on chromosome 4, 3,181,000 to 3,195,000 (Fig. 5).

Group A strains exhibited a shared pattern of deletions and mutations within their gene clusters. The CPA gene cluster pattern depended on the aflatoxin gene cluster pattern. The strain that had a deletion in the aflatoxin gene cluster had a large deletion in the CPA gene cluster, while other strains showed consistent mutation patterns. For the ditryptophenaline cluster, Group A strains shared a similar mutation pattern.

Group B showed similar mutation patterns with part of the Group A strains in the aflatoxin and CPA gene clusters, but they had intact ditryptophenaline gene clusters.

Most of the Group C strains had intact three gene clusters. Conversely, three aflatoxigenic strains (AR028, SD016, and SL015) exhibited unique mutation patterns in the aflatoxin and CPA gene clusters. This mutation pattern is odd compared to Group A and B mutation patterns.

Group D strains presented a distinctive pattern with large deletions in the aflatoxin cluster and complete deletion in the CPA cluster, maintaining an intact Ditryptophenaline gene cluster.

In Group E, KRM strains and Afla-guard strains showed distinct features. In contrast to other strains, where the CPA gene cluster pattern depends on the aflatoxin gene cluster's state, KRM strains uniquely exhibited a deleted aflatoxin gene cluster while retaining the CPA gene cluster. Afla-guard strains had a whole deletion in the aflatoxin and CPA gene clusters. However, these Group E strains had an intact ditryptophenaline gene cluster similar to the others. No group-specific characteristics were observed among the groups concerning Aspergillic acid and Aflatrem gene clusters.

3.4. MALDI TOF/MS patterning

Comparing patterns of the peaks of Group C and the non-C groups (A, B and E; A. oryzae groups), the two groups were distinctly separated, while non-C groups were not distinctly separated from each other (Supplementary Fig. 1).

Group C and the non-C groups exhibited several significantly different loci; 3100–3200 m/z, 6200–6500 m/z, and 12700–12900 m/z showed significant differences (Fig. 6B). When comparing the non-C groups, no distinct loci were found that significantly differentiated between the groups in overall ranges (Supplementary Fig. 2). Most of the Group C strains had significantly higher peaks in the 6355 to 6385 m/z and 12795 to 12810 m/z ranges, while most of the non-C groups had high peaks at 3145 m/z, 6285 to 6295 m/z, and 12730 m/z.

The phylogenetic analysis and genome structure investigation of A. oryzae/flavus strains from fermentation and wild environment in Korea, compared with globally reported strains, provided significant insights into the genetic diversity and evolutionary pathways of these fungi. The phylogenetic tree, PCA and ADMIXTURE analysis offer a vivid illustration of the genetic distances and relationships among the strains, revealing a broad spectrum of evolutionary divergence within five distinct groups (A to E). In particular, Groups A and B demonstrate closer genetic affiliations, suggesting a shared evolutionary pathway that may be rooted in their non-aflatoxigenic nature and potential adaptations to industrial or natural environments. Conversely, Group E, characterized by its distinct genetic makeup, stands out as the most genetically divergent cluster. This may indicate unique evolutionary pressures or historical genetic isolation that have shaped their current genomic constitution. Group D exhibits an interesting feature since it has a greater genetic distance from Groups A and B than from Group C, yet PCA analysis showed that it aligns more closely with Groups A and B.

Considering distribution of Korean strains, industrial strains (KRI), excluding KRI2, clustered within Group A, specifically aligning with JPI Clade C strains. The narrow genetic distances among these industrial strains suggest recent differentiation driven by fermentation functionalities, resulting in their separation as individual strains.

Korean Meju-originated strains (KRM) exhibited broader distribution across Groups A, B, and E. Some strains were closely matched to industrial strains, likely due to spore dispersal from industrial environments to traditional fermentation settings. Other Meju strains in Group A, B and E showed distinct genetic positions with industrial strains. Notably, two Meju strains in Group E showed genetic similarity with the biocontrol agent, Afla-guard but differed in aflatoxin and CPA gene clusters, suggesting potential genetic recombination events with wild strains. The distribution of Meju strains(KRM) was similar with that of non-aflatoxigenic strain from wild conditions (KRWO). This supports the nothion that Meju fermentation involves environmental strains without artificial starter inoculation, resulting in weak selective pressure.

Non-aflatoxigenic strains from Korean wild environments (KRWO) were found in Groups A and B. Group A wild strains exhibited genetic distinctions from JPI strains, suggesting different evolutionary pressures. The lower genetic diversity among these wild strains, compared to industrial strains, may indicate fewer evolutionary changes in natural settings. In contrast, industrial environments likely experienced more frequent selective evolution events.

All aflatoxigenic Korean strains (KRWF) were categorized in Group C. Especially strains AR028, SD016, and SL015 had notable genetic characteristics highlighted in previous studies [35]. PCR analysis patterns of the norB/cypA cluster in these strains were similar to that of A. parasiticus. Genome-wide clustering confirmed their placement within Group C, yet their aflatoxin cluster mutation patterns were notably distinct from NRRL 3357 and other Group C strains. Despite these genomic differences, MALDI-TOF MS results closely matched NRRL 3357 and other Group C strains.

Following Aflatoxin mutation patterns, this study reveals that there are points of concordance with previous research as well as advancements in understanding these patterns. Comparing the strains, Afla-guard (NRRL 21882), which had a complete deletion of the aflatoxin gene cluster, and strain 14160, which exhibited deletion starting from the OmtA gene showed the same results. Strains BP2-1, TK-10, TK24, WRRL1519, and NRRL35739 showed deletions beginning from the norA gene. These results are also consistent with previous findings.

Additionally, strains 3.042, TK-5, TK-59, and SU-16 were described as having mostly intact gene clusters with partial deletions starting from the AflT gene. Group C strains (A9, CA14, NRRL 3357, VCG1, Washington T5, E1404, E1445) were reported to have partial deletions starting from the CypA gene [10]. In this study, non-aflatoxigenic strains, including those in Group C, frequently exhibited additional deletions of the cypA gene compared to NRRL 3357. This suggests that the additional deletion in cypA may significantly impact aflatoxin production. This finding also aligns with previous research that classified norB-cypA PCR patterns into type 1 (short read type) and type 2 (long read type), with aflatoxigenic strains being type 2 and non-aflatoxigenic strains being type 1 [35].

Across Groups A and B, strains with high mutation rates or high deletion rates are mixed and each of them shares a similar mutation pattern. But they don’t share similar mutation patterns with Group C strains. This suggests significant genetic exchange through interbreeding between Groups A and B during domestication, while the two groups appear to be reproductively isolated from the aflatoxigenic group, indicating speciation. Therefore, these strains with high mutation rates or high deletion rates are considered safe since they have less potential for producing mycotoxins and are unlikely to form hybrids with aflatoxigenic strains.

Group C mostly consists of aflatoxigenic strains (KRWF and NF), but there are some exceptional non-aflatoxigenic strains (2017 Washington-T2, A1, AF36 and K54A). Among them, AF36 is even known as biocontrol agent to reduce aflatoxin contamination [36]. This may indicate a complex genetic landscape that does not strictly correlate with aflatoxin production. However, the presence of intact aflatoxin and cyclopiazonic acid gene clusters in these non-aflatoxigenic strains suggests a latent potential for mycotoxin production under certain conditions, raising important considerations for food safety and strain utilization in industrial applications..

The evolutionary direction of Group A, characterized by unique mutations in the ditryptophenaline gene cluster, indicates a potential adaptive response to industrial fermentation processes, possibly due to selective pressures to mitigate negative impacts on food products. Although the toxicity or side effects of ditryptophenaline have not been thoroughly studied, the uniform occurrence of mutations in this cluster among Group A strains implies that such mutations may confer a selective advantage in the context of fermented food production, possibly due to reduced detrimental effects on product quality [7].

By incorporating MALDI-TOF MS data, we observe significant differentiation, particularly between Group C and the non-C groups (A, B, and E), with unique proteomic features corresponding to their genomic distinctions. Group C exhibited the unique features in the ranges of 3100–3200 m/z, 6200–6500 m/z, and 12700–12900 m/z. A database of various proteins specific for A. flavus and A. oryzae in UniProtKB (www.uniprot.org) suggests that these peaks may correspond to significant fungal proteins. For instance, peaks within the 6200 to 6500 m/z range correspond to metabolic enzyme proteins like 3-Hydroxyanthranilate 3,4-Dioxygenase (3HAO) and Short-chain dehydrogenase. Peaks within the 12700 to 12900 m/z range may correspond to proteins such as AflF, also known as norB (primary accession number: A0A7U2MNK6, Mass: 12855 Da). This protein is one of the enzymes included in the aflatoxin gene cluster. No characterized proteins within the 3100 to 3200 m/z range were identified in the database. These specific proteins play crucial roles in cellular metabolism, suggesting their potential as reliable biomarkers for differentiating between groups.

Previous attempts to differentiate A. oryzae from A. flavus have been numerous. A previous study showed that targeting the Cyp51A gene could provide differentiation, but the limited strain diversity hindered its broad applicability [37]. There were also attempts to analyze aflatoxin and cyclopiazonic acid gene clusters to see the difference between two species, and although some differentiation was achieved, it was not definitive [3, 38]. There have also been attempts to distinguish the two species through profiling of CAZyme genes and secondary metabolite biosynthesis gene clusters, but the ambiguous similarities between the two species disturbed clear differentiation [10].

In this study, a comprehensive approach was employed, combining whole-genome SNP-based population structure analysis, detailed secondary metabolite gene cluster variation analysis, and MALDI-TOF MS profiling.

Group C formed a distinct cluster in genomic analysis, separating it from other groups. Similarly, MALDI-TOF MS analysis distinguished Group C from other groups, displaying unique peaks. Furthermore, group C strains have comparatively intact gene clusters of aflatoxin and cyclopiazonic acid and produced aflatoxin B. Given these characteristics, Group C aligns with traits traditionally associated with A. flavus. Therefore, this study proposes classifying Group C as A. flavus and the non-C groups as A. oryzae.

This study provides a comprehensive genomic and proteomic characterization of Korean A. oryzae/flavus complex strains compared to global strains. The analysis highlights the genetic diversity and distinct evolutionary paths of these strains, with particular attention to their safety for food industry applications. Genomic analysis divided the A. oryzae/flavus strains into five groups, with Group C showing relatively intact aflatoxin and cyclopiazonic acid (CPA) gene clusters, suggesting a high potential for producing both metabolites. Furthermore, unique proteomic patterns identified through MALDI-TOF MS provide reliable biomarkers for distinguishing Group C aflatoxigenic strains. Characteristics of group C were well matched with those of A. flavus and the other groups with A. oryzae. These findings establish refined criteria for differentiating A. oryzae from A. flavus, contributing to their safe utilization and better understanding of their genetic dynamics.

CPA

Cyclopiazonic acid

KRI

Korean Industrial non-aflatoxigenic strains

KRM

Korean non-aflatoxigenic strains from meju

KRWO

Korean non-aflatoxigenic strains from wild conditions

KRWF

Korean aflatoxigenic strains from wild conditions

JPI

Japanese Industrial strains

NCBI non-aflatoxigenic strains

NCBI aflatoxigenic strains

Afla-guard non-aflatoxigenic strains.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Not applicable.

Competing interests

Not applicable.

Funding

This work is supported by the grant from National Institute of Agricultural Sciences, Rural Development Administration, Republic of Korea (PJ017286), and the Ministry of Science, ICT (MSIT) of Korea (2021M3H9A1081101).

Authors' contributions

D.H. Kim wrote the main manuscript text and prepared figures 1-5. S.B. Hong and K.T. Kim provided revisions to the contents of main manuscript. D.C. Kim and D.G. Seo prepared figure 6. S.H. Lee reviewed and corrected the manuscript’s grammar and language.

Acknowledgements

Not applicable.

Machida M, et al. Genome sequencing and analysis of Aspergillus oryzae. Nature. 2005;438(7071):1157–61.
Machida M, Yamada O, Gomi K. Genomics of Aspergillus oryzae: learning from the history of Koji mold and exploration of its future. DNA Res. 2008;15(4):173–83.
Gibbons JG, et al. The evolutionary imprint of domestication on genome variation and function of the filamentous fungus Aspergillus oryzae. Curr Biol. 2012;22(15):1403–9.
Alshannaq AF, et al. Controlling aflatoxin contamination and propagation of Aspergillus flavus by a soy-fermenting Aspergillus oryzae strain. Sci Rep. 2018;8(1):16871.
Watarai N, et al. Evolution of Aspergillus oryzae before and after domestication inferred by large-scale comparative genomic analysis. DNA Res. 2019;26(6):465–72.
Kjærbølling I, et al. A comparative genomics study of 23 Aspergillus species from section Flavi. Nat Commun. 2020;11(1):1106.
Rank C, et al. Comparative chemistry of Aspergillus oryzae (RIB40) and A. flavus (NRRL 3357). Metabolites. 2012;2(1):39–56.
Medina A, Rodriguez A, Magan N. Effect of climate change on Aspergillus flavus and aflatoxin B1 production. Front Microbiol. 2014;5:348.
Sun H, et al. Safety evaluation and comparative genomics analysis of the industrial strain Aspergillus flavus SU-16 used for huangjiu brewing. Int J Food Microbiol. 2022;380:109859.
Han DM, et al. Comparative pangenome analysis of Aspergillus flavus and Aspergillus oryzae reveals their phylogenetic, genomic, and metabolic homogeneity. Food Microbiol. 2024;119:104435.
Dumas E, et al. Independent domestication events in the blue-cheese fungus Penicillium roqueforti. Mol Ecol. 2020;29(14):2639–60.
Kim E, et al. Differentiation between Weissella cibaria and Weissella confusa using machine-learning-combined MALDI-TOF MS. Int J Mol Sci. 2023;24(13):11009.
Hedayati MT, et al. Discrimination of aspergillus flavus from Aspergillus oryzae by matrix-assisted laser desorption/ionisation time‐of‐flight (MALDI‐TOF) mass spectrometry. Mycoses. 2019;62(12):1182–8.
Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20.
Fountain JC, et al. Draft genome sequences of one Aspergillus parasiticus isolate and nine Aspergillus flavus isolates with varying stress tolerance and aflatoxin production. Microbiol Resource Announcements. 2020;9(37). p. 10.1128/mra. 00478 – 20.
Chacón-Vargas K, et al. Comparison of two Aspergillus oryzae genomes from different clades reveals independent evolution of alpha-amylase duplication, variation in secondary metabolism genes, and differences in primary metabolism. Front Microbiol. 2021;12:691296.
Pennerman KK, et al. Aspergillus flavus NRRL 35739, a poor biocontrol agent, may have increased relative expression of stress response genes. J Fungi. 2019;5(2):53.
Chang PK. Genome-wide nucleotide variation distinguishes Aspergillus flavus from Aspergillus oryzae and helps to reveal origins of atoxigenic A. flavus biocontrol strains. J Appl Microbiol. 2019;127(5):1511–20.
Schamann A, Geisen R, Schmidt-Heydt M. Draft genome sequence of an aflatoxin-producing Aspergillus flavus strain isolated from food. Microbiol Resource Announcements. 2022;11(2):e00894–21.
Arias RS, et al. Sixteen draft genome sequences representing the genetic diversity of Aspergillus flavus and Aspergillus parasiticus colonizing peanut seeds in Ethiopia. Microbiol Resource Announcements. 2020;9(30). p. 10.1128/mra. 00591 – 20.
Hua SST, et al. Characterization of aflatoxigenic and non-aflatoxigenic Aspergillus flavus isolates from pistachio. Mycotoxin Res. 2012;28:67–75.
Li H, Durbin R. Fast and accurate long-read alignment with Burrows–Wheeler transform. Bioinformatics. 2010;26(5):589–95.
Li H, et al. The sequence alignment/map format and SAMtools. Bioinformatics. 2009;25(16):2078–9.
Van der Auwera GA et al. From FastQ data to high-confidence variant calls: the genome analysis toolkit best practices pipeline. Current protocols in bioinformatics, 2013. 43(1): p. 11.10. 1-11.10. 33.
Rempel A, Wittler R. SANS serif: alignment-free, whole-genome-based phylogenetic reconstruction. Bioinformatics. 2021;37(24):4868–70.
Chang CC et al. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience, 2015. 4(1): p. s13742-015-0047-8.
Yang J, et al. GCTA: a tool for genome-wide complex trait analysis. Am J Hum Genet. 2011;88(1):76–82.
Alexander DH, Novembre J, Lange K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 2009;19(9):1655–64.
Yu J, et al. Clustered pathway genes in aflatoxin biosynthesis. Appl Environ Microbiol. 2004;70(3):1253–62.
Amaike S, Keller NP. Aspergillus flavus. Annu Rev Phytopathol. 2011;49:107–33.
Caceres I, et al. Aflatoxin biosynthesis and genetic regulation: A review. Toxins. 2020;12(3):150.
Chang P-K, Ehrlich KC, Fujii I. Cyclopiazonic acid biosynthesis of Aspergillus flavus and Aspergillus oryzae. Toxins. 2009;1(2):74–99.
Kishimoto S, et al. Evaluation of biosynthetic pathway and engineered biosynthesis of alkaloids. Molecules. 2016;21(8):1078.
Alexander DH, Lange K. Enhancements to the ADMIXTURE algorithm for individual ancestry estimation. BMC Bioinformatics. 2011;12:1–6.
Hong S-B, et al. The proportion of non-aflatoxigenic strains of the Aspergillus flavus/oryzae complex from meju by analyses of the aflatoxin biosynthetic genes. J Microbiol. 2013;51:766–72.
Doster MA, Cotty PJ, Michailides TJ. Evaluation of the atoxigenic Aspergillus flavus strain AF36 in pistachio orchards. Plant Dis. 2014;98(7):948–56.
Nargesi S, et al. Differentiation of Aspergillus flavus from Aspergillus oryzae targeting the cyp51A gene. Pathogens. 2021;10(10):1279.
Toyotome T, et al. Comparative genome analysis of Aspergillus flavus clinically isolated in Japan. DNA Res. 2019;26(1):95–103.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
09 Aug, 2024
Editor assigned by journal
08 Aug, 2024
Submission checks completed at journal
08 Aug, 2024
First submitted to journal
06 Aug, 2024

You are reading this latest preprint version

Genome Diversity, Population Structure and MALDI-TOF MS Profiling of Aspergillus oryzae/flavus Strains from Fermentation and Wild Environments

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials & Methods

2.1. Collection of Genomes of Korean and Global A. oryzae/flavus Strains

2.2. Single Nucleotide Polymorphisms (SNPs) calling and population genetic analysis

2.3. Genotypic Analysis of Secondary Metabolites Gene Cluster and Design of Diagnostic Markers

2.4. MALDI TOF metabolites pattern analysis

3. Results

3.1. Phylogenetic relationship of A. oryzae/flavus strains

3.2. Genome and Population Structure Analysis of A. oryzae/flavus

3.3. Secondary Metabolite Gene Cluster Analysis

3.4. MALDI TOF/MS patterning

4. Discussion

5. Conclusion

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1