Aspergillus parasiticus SU-1 Genome Sequence, Predicted Chromosome Structure, and Comparative Gene Expression under Aflatoxin-Inducing Conditions: Evidence that Differential Expression Contributes to Species Phenotype

Abstract

The filamentous fungi Aspergillus parasiticus and Aspergillus flavus produce the carcinogenic secondary metabolite aflatoxin on susceptible crops. These species differ in the quantity of aflatoxins B₁, B₂, G₁, and G₂ produced in culture, in the ability to produce the mycotoxin cyclopiazonic acid, and in morphology of mycelia and conidiospores. To understand the genetic basis for differences in biochemistry and morphology, we conducted next-generation sequence (NGS) analysis of the A. parasiticus strain SU-1 genome and comparative gene expression (RNA sequence analysis [RNA Seq]) analysis of A. parasiticus SU-1 and A. flavus strain NRRL 3357 (3357) grown under aflatoxin-inducing and -noninducing culture conditions. Although A. parasiticus SU-1 and A. flavus 3357 are highly similar in genome structure and gene organization, we observed differences in the presence of specific mycotoxin gene clusters and differential expression of specific mycotoxin genes and gene clusters that help explain differences in the type and quantity of mycotoxins synthesized. Using computer-aided analysis of secondary metabolite clusters (antiSMASH), we demonstrated that A. parasiticus SU-1 and A. flavus 3357 may carry up to 93 secondary metabolite gene clusters, and surprisingly, up to 10% of the genome appears to be dedicated to secondary metabolite synthesis. The data also suggest that fungus-specific zinc binuclear cluster (C6) transcription factors play an important role in regulation of secondary metabolite cluster expression. Finally, we identified uniquely expressed genes in A. parasiticus SU-1 that encode C6 transcription factors and genes involved in secondary metabolism and stress response/cellular defense. Future work will focus on these differentially expressed A. parasiticus SU-1 loci to reveal their role in determining distinct species characteristics.

INTRODUCTION

Aspergillus parasiticus and Aspergillus flavus are the predominant aflatoxin producers on susceptible crops (1). The A. flavus NRRL 3357 (3357) complete genome sequence is available (http://www.aspergillusflavus.org/genomics/), and the sequence and location of the aflatoxin biosynthetic gene cluster were determined as part of this analysis. The nucleotide sequence of the A. parasiticus SU-1 aflatoxin biosynthetic gene cluster is also available (2), and A. flavus 3357 and A. parasiticus SU-1 exhibit a high degree of sequence identity between the aflatoxin genes as well as between other genes that have been sequenced in both organisms.

Although it is clear that A. flavus and A. parasiticus are closely related species, several phenotypic characteristics have been used to differentiate these species; these phenotypic characteristics include the quantity and type of mycotoxins synthesized in culture and on plants. For example, A. flavus strains synthesize predominantly B aflatoxins, while A. parasiticus strains produce both B- and G-type aflatoxins (1, 3). Less than 50% of A. flavus isolates are reported to produce aflatoxins, while almost all strains of A. parasiticus are toxigenic (4). A. parasiticus strains generally synthesize greater quantities of aflatoxin than A. flavus strains do (1, 3). A. flavus strains are also reported to synthesize cyclopiazonic acid (CPA), while A. parasiticus strains are not known to synthesize this mycotoxin (1).

A. parasiticus and A. flavus also exhibit morphological differences, and gross mycelial structure and conidiospore morphology have been used to help assign species designations to isolates in the genus Aspergillus (5). A. flavus and A. parasiticus can grow and produce aflatoxin on a wide variety of susceptible crops. A. flavus is the most widely reported fungus associated with food, and it appears to have a broader host range because it is strongly associated with contamination of corn, cotton, peanuts, and tree nuts, while A. parasiticus is primarily associated with contamination of peanuts (4). Both species share an ecological niche in the soil, but it is not clear how well these organisms compete there.

To identify the genetic basis for reported differences in species morphology and biochemistry, we conducted Illumina next-generation sequence (NGS) analysis on the A. parasiticus SU-1 genome. We compared the A. parasiticus SU-1 genome sequence and predicted chromosome structure with the available A. flavus 3357 genome and compared expression (RNA sequence analysis [RNA Seq]) of all detectable A. parasiticus SU-1 and A. flavus 3357 genes under identical aflatoxin-inducing (YES growth medium, 40 h of growth; standard conditions, see Materials and Methods) and aflatoxin-noninducing (YEP medium, 40 h) growth conditions. We observed that the genome sequences of A. parasiticus SU-1 and A. flavus 3357 exhibit a high degree of identity, and the chromosome structure and gene distribution are predicted to be similar. Despite these similarities, we observed significant differences in the presence of specific secondary metabolism genes and entire gene clusters as well as differential expression of secondary metabolism and other genes between the two species that may contribute to species differences in phenotype. In addition, we identified three sets of uniquely expressed genes in A. parasiticus that encode fungus-specific zinc binuclear cluster (C6) transcription factors as well as enzymes involved in secondary metabolism and stress response/cellular defense. Identification of these differentially expressed genes prompts further study of the genetic basis for unique phenotypes observed in these closely related fungal species.

MATERIALS AND METHODS

Microorganisms and culture conditions.

Aspergillus parasiticus strain SU-1 (ATCC 56775) and A. flavus strain NRRL 3357 (3357) were used as the wild-type aflatoxin-producing strains. YES liquid medium (2% yeast extract, 6% sucrose [pH 5.8]) (6) was used as an aflatoxin-inducing rich growth medium, and YEP liquid medium (2% yeast extract, 6% peptone [pH 5.8]) (7) was used as an aflatoxin-noninducing medium. An aliquot of frozen conidiospores from each strain was used to inoculate 100 ml of these growth media (250-ml Erlenmeyer flask carrying 5- or 6-mm glass beads) at 10⁴ spores per ml, and these cultures were incubated for 40 h in the dark at 30°C with shaking (150 rpm) as described previously (standard growth conditions) (7).

DNA preparation and Illumina NGS analysis.

Genomic DNA was prepared from A. parasiticus SU-1 grown for 44 h in YES medium under standard conditions using the method of Cihlar and Sypherd (8) as modified by Horng et al. (9). Genomic DNA was subjected to Illumina next-generation sequence (NGS) analysis by Functional Biosciences (Madison, WI), and bioinformatics analysis was conducted by ContigExpress, LLC (New York, NY). Quality control checks on the raw sequencing data (80 million reads in the forward and reverse directions) were performed using FastQC version 0.10.1. K-mer error correction was performed on the raw reads using Quake version 0.3.4. Paired-end reads were extracted from the corrected read pool, and the remaining reads were deposited as single-end reads. Both paired-end and single-end reads were used in de novo genome assembly. SOAPDenovo version 1.05 was used to perform assembly of the error-corrected reads. Standard parameters for paired-end reads were used. The K-mer that generated scaffold sequences with a maximal N₅₀ was 67. The resulting scaffolds were gap filled using corrected paired-end reads. The final scaffold sequences consisted of 375 gap-filled scaffold sequences larger than 1 kb (see A. parasiticus FASTA scaffolds in the supplemental file in the supplemental material).

Comparison of chromosome structure and gene distribution in A. parasiticus SU-1 and A. flavus 3357.

The 31 largest scaffolds from the A. parasiticus SU-1 de novo genome sequence assembly were compared to the 16 largest A. flavus 3357 scaffolds that composed the arms of the eight chromosomes in this filamentous fungus (http://www.aspergillusflavus.org/genomics/). The relative position of each of the 31 A. parasiticus SU-1 scaffolds was assigned by direct sequence comparison using the 16 largest A. flavus 3357 scaffolds as a reference. These position assignments were supported by examination of the relative locations of 43 secondary metabolite gene clusters that were shared by A. flavus 3357 and A. parasiticus SU-1. The locations of A. parasiticus SU-1 open reading frames on A. parasiticus SU-1 scaffolds and chromosomes also were completed based on direct sequence comparison of the largest A. flavus 3357 and A. parasiticus SU-1 scaffolds.

antiSMASH (antibiotics and secondary metabolite analysis shell) software (10) was used to identify secondary metabolite loci within the A. flavus 3357 and A. parasiticus SU-1 genomes. FASTA formatted files of the A. flavus 3357 genome sequence (NCBI accession number AAIH00000000) and the A. parasiticus SU-1 de novo genome assembly were uploaded onto the web version of antiSMASH (http://antismash.secondarymetabolites.org/) with default settings. antiSMASH output displays a list of identified clusters based on the chemical backbones of their predicted products.

The predicted function of the uniquely expressed genes was determined by ContigExpress, LLC, New York, NY. Raw FASTA reads of 213 uniquely expressed transcripts in A. parasiticus SU-1 were translated in six frames and submitted as BLAST queries using NCBI nonredundant protein sequences. Function and Gene Ontology (GO) assignment was conducted using BLAST2GO. A subsequent BLAST of the translated transcript sequences was compared against a Kyoto Encyclopedia of Genes and Genomes (KEGG) database and assigned KEGG gene identification (ID) and pathway analysis using KOBAS. Function, GO, and functional pathway analysis were compiled in an Excel file.

RNA preparation and RNA Seq analysis.

Total RNA was isolated from two biological replicates of A. parasiticus SU-1 and two biological replicates of A. flavus NRRL 3357 grown under standard conditions in YES medium for 40 h using the TRIzol method (TRIzol reagent; Invitrogen, Carlsbad, CA) as described by the manufacturer. Total RNA was also isolated from a single sample of A. parasiticus SU-1 grown in YEP medium under standard conditions for 40 h. The integrity and purity of the RNA samples were assessed using an Agilent 2200 TapeStation bioanalyzer system (Agilent, Santa Clara, CA). All samples had an RNA integrity number (RIN) of 9.0 or higher with an optical density at 260 or 280 nm (OD_260/280) between 1.78 and 1.80. Illumina RNA sequence (RNA Seq) analysis was conducted on purified total RNA samples by Otogenetics (Norcross, GA), and bioinformatics analysis of the resulting data was conducted by ContigExpress, LLC (New York, NY).

cDNA library construction was conducted as follows. One hundred nanograms of total RNA was reverse transcribed into cDNA using a Clontech SmartPCR cDNA kit (Clontech, Mountain View, CA). A Fragment Analyzer (Advanced Analytical Technologies, Inc. [AATI], Ames, IA) and Qubit (Life Technologies, Grand Island, NY) were used to assess cDNA quality. Restriction enzyme digestion end repair was conducted by removing adaptor sequences, and fragmentation of the resulting cDNA was conducted using a Covaris M220 focused ultrasonicator (Woburn, MA). Fragmented cDNA was subjected to Illumina library preparation with NEBNext-based reagents (New England BioLabs [NEB], Ipswich, MA). The quality, quantity, and size distribution of Illumina libraries were determined using an Agilent Bioanalyzer 2100. cDNA clusters were generated on a cBot automated station, and the Illumina libraries were submitted for Illumina HiSeq 2000 100-bp paired-end sequencing (Illumina, San Diego, CA).

Data analysis workflow was developed by ContigExpress, LLC (New York, NY). FASTQC (Babraham Institute, Cambridge, United Kingdom) was used to assess quality of paired-end 100-bp Illumina raw sequence reads. The numbers of raw reads per RNA sample were as follows: A. parasiticus SU-1 in YES medium, replicate 1 (35,870,008); SU-1 in YES medium, replicate 2 (23,352,926); A. flavus 3357 in YES medium, replicate 1 (21,887, 60); 3357 in YES medium, replicate 2 (28,694,452); SU-1 in YEP medium (22,730,224). Processed reads were then mapped to the de novo scaffold sequences of the A. parasiticus SU-1 genome with paired-end configuration using Bowtie2, version 2.0.0-beta7 (11). The A. flavus 3357 genome (http://www.aspergillusflavus.org/genomics/) was used to obtain A. flavus cDNA gene designations which were compared to the scaffold sequences using BLAST, version 2.2.24+ (12). The matched scaffold region of the top hit of each A. flavus gene, when available, was used to annotate that region with an A. flavus gene ID. The annotation information was assembled in the GTF format for downstream analysis. The mapping files of three A. parasiticus samples and two A. flavus samples were used to assemble the master gene annotations. The BLAST-based gene annotation was submitted to Cufflinks (version 2.0.2) as the reference annotation (13). Differential gene expression analysis was performed with the master gene annotations using Cufflinks.

Identification and function of “uniquely expressed loci” in A. parasiticus.

We compared the level of gene expression in the two A. flavus 3357 biological replicates grown for 40 h on YES medium under standard conditions versus the two A. parasiticus SU-1 biological replicates grown for 40 h under the same conditions. We compiled a list of 1,409 genes that exhibited significant levels of differential expression between the two fungal species grown for 40 h on YES medium (q value < 0.05). Of these 1,409 genes, 318 were not identified in the A. flavus 3357 genome (http://www.aspergillusflavus.org/genomics/). We hypothesized that either these genes were truly absent in A. flavus 3357, or they were omitted due to gaps in coverage in the A. flavus 3357 genome sequence. We also observed that of the 318 genes not present in the A. flavus 3357 genome, 213 were expressed at significant levels in A. parasiticus SU-1, but the transcript was not detected in A. flavus 3357. Because these 213 genes and their transcript were absent in A. flavus 3357 but present in A. parasiticus SU-1, we propose that they represent genes that are uniquely expressed in A. parasiticus SU-1 (compared to A. flavus 3357). For the remaining 105 genes, transcripts were detected in A. flavus 3357, suggesting that they were not identified previously due to gaps in construction of the genome sequence.

Nucleotide sequence accession number.

The A. parasiticus SU-1 scaffold sequences have been submitted to NCBI and assigned accession number JMUG00000000.

RESULTS

A. parasiticus SU-1 genome sequence and predicted chromosome structure.

As part of the bioinformatics analysis, contigs from Illumina sequence analysis of the A. parasiticus SU-1 genome were assembled into 375 scaffolds greater than 1 kb in size (see Table S1 in the supplemental material). The resulting scaffolds were then listed by decreasing length. The top 31 A. parasiticus SU-1 scaffolds by length represented approximately 95% of the predicted SU-1 genome (Table 1). The locations of all 375 A. parasiticus SU-1 scaffolds were assigned based on direct sequence comparison with 16 A. flavus 3357 scaffolds presented in the A. flavus 3357 genome browser (http://www.aspergillusflavus.org/genomics/) (see Table S2 in the supplemental material). The locations of the top 31 A. parasiticus SU-1 scaffolds by length were assigned to A. flavus 3357 chromosomes by a similar analysis, and these data were used to construct a schematic of the A. parasiticus SU-1 genome (Fig. 1A). We observed that the predicted chromosome structures of the A. parasiticus SU-1 and A. flavus 3357 genomes are quite similar (Fig. 1A and B). The A. parasiticus SU-1 genome is predicted to consist of 8 chromosomes that have sizes similar to those of the 8 homologous A. flavus 3357 chromosomes. A total of 13,290 A. parasiticus SU-1 genes were identified based on the A. flavus 3357 genome annotation (carries 13,487 genes), and the relative locations of these loci were assigned to A. parasiticus SU-1 scaffolds and chromosomes by direct sequence comparison to the homolog in A. flavus 3357 scaffolds and chromosomes (see Table S3 in the supplemental material).

TABLE 1.

Top 31 A. parasiticus SU-1 scaffolds by length

A. parasiticus scaffold no.	Scaffold length (bp)	Scaffold locationa
35	3,001,269	C2R
222	2,822,244	C4L
310	2,638,474	C7L
77	2,214,935	C3R
311	2,173,408	C2L
32	2,057,647	C1R
44	1,914,451	C4R
26	1,893,530	C1L
30	1,776,511	C6R
139	1,659,146	C5R
129	1,552,207	C3L
64	1,506,959	C1R
18	1,451,838	C5L
69	1,326,957	C8R
29	1,267,361	C2R
123	1,129,355	C3L
25	1,079,114	C5L
309	793,882	C8L
8	646,144	C8R
22	639,763	C1R
17	564,883	C6L
28	492,974	C6L
46	434,019	C6R
11	433,510	C8L
308	340,505	C6L
126	338,060	C6L
208	319,820	C7R
253	225,885	C5R
215	179,257	C6L
307	169,918	C1R
76	154,073	C3R

^aThe location of the scaffold on the chromosome is shown as follows: the chromosome number is shown first and then the arm (e.g., C2R is chromosome 2, right arm, and C4L is chromosome 4, left arm).

FIG 1.

Predicted chromosome structure of A. parasiticus SU-1 and A. flavus 3357. (A) The A. parasiticus (AP) SU-1 genome sequence was assembled into scaffolds, and the positions of the largest 31 scaffolds (S) were located on the left arm (L) or right arm (R) of 8 chromosomes as described in Materials and Methods. The number following the scaffold number indicates its length in base pairs. The chromosome number appears in bold type to the right of the right arm of each chromosome, and the size of the entire chromosome (megabase pairs [MBp]) is included in parentheses next to the chromosome number. The relative location and size (in base pairs) of the largest 31 A. parasiticus SU-1 scaffolds (i.e., S26, S22, S64) appears within each chromosome arm, and the size of each arm (MBp) appears near the right-hand end in parentheses. The locations of 44 secondary metabolism gene clusters are indicated by small blue ellipses (filled with red) placed just below the chromosome arms, and the cluster numbers correspond to the gene cluster designations in the A. flavus genome browser cited in the text. (B) Predicted A. flavus (AF) 3357 chromosome structure based on the genome browser. Eleven secondary metabolite clusters present in A. flavus 3357 (clusters 4, 22, 28, 33, 36, 38, 40, 45, 48, 49, and 55) were not identified in A. parasiticus SU-1 based on manual analysis for a single key cluster gene. Cluster 36 was also located in different chromosomes in strains SU-1 and 3357.

A. parasiticus SU-1 genes were provided A. flavus 3357 gene designations based on identity, and these genes were located on the A. parasiticus SU-1 scaffolds and chromosomes using the A. flavus 3357 genome as a template. We observed that 96% of A. flavus 3357 open reading frames (ORFs) have at least one homolog in the A. flavus 3357 genome, and 4% of the A. flavus 3357 ORFs have no detectable homolog in the A. parasiticus SU-1 genome (Fig. 2A). A. parasiticus SU-1 and A. flavus 3357 share greater than 90% sequence identity over greater than 90% of genome sequence (Fig. 2B).

FIG 2.

Analysis of genome identity between A. parasiticus SU-1 and A. flavus 3357. (A) Percentage of A. flavus 3357 open reading frames with at least one homolog in A. parasiticus SU-1. (B) Distribution of identity over the A. parasiticus SU-1 genome sequence. The percentage in the pie chart represents the fraction of the genome that carries the identity level (indicated by color).

Fifty-five secondary metabolite gene clusters were identified and located on A. flavus 3357 chromosomes as part of the original sequence analysis of the A. flavus 3357 genome. We screened for these secondary metabolism gene clusters in the A. parasiticus SU-1 genome by BLAST analysis using one key gene in each cluster as the query sequence and the A. parasiticus scaffolds as the subject sequence. We detected the key gene from 44 of the A. flavus 3357 secondary metabolite gene clusters in A. parasiticus SU-1, and these clusters were found in similar locations on the A. parasiticus SU-1 chromosomes (Fig. 1 and Table 2). The exception is cluster 34, which appears to be located on the left arm of chromosome 2 (C2L) rather than the left arm of chromosome 6 (C6L). On the basis of this labor-intensive manual analysis, we noted that A. parasiticus SU-1 lacks 11 of the secondary metabolite clusters observed in A. flavus 3357, including clusters 4, 22, 28, 33, 36, 38, 40, 45, 48, 49, and cluster 55, which mediates synthesis of cyclopiazonic acid (CPA) (Table 2).

TABLE 2.

Secondary metabolism gene clusters in the A. parasiticus SU-1 genome screened by BLAST analysis^a

SM cluster no.b	AFLA gene no.c	Location of the SM gene cluster		Chromosomal location of the SM gene clusterd	Location on A. flavus scaffold	Gene size (bp)	Location on A. parasiticus scaffolde	Level of expressionf
		Scaffold no. in A. flavus	Scaffold no. in A. parasiticus					SU-1 YES/YEP	3357 YES/SU-1 YEP	3357/SU-1 YES
18	NRPS 5427	83	26	C1L	1,804,577	1,951	110,065	0, 0 no	0, 0 no	0, 0 no
17	PKS 5287	83	26	C1L	1,688,441	9,065	223,202	87,113 no	83, 113 no	91, 83 no
29	DMAT 8408	72	64	C1R	3,179,694	2,056	962,759	0.1, 0 no	1.6, 0 yes	1.6, 0.1 no*
28	NRPS 8248	72	None	C1R	2,730,816	1,252	None
27	PKS 8215	72	64	C1R	2,646,293	6,545	426,236	0, 0 no	0, 0 no	0, 0 no
26	NRPS 7940	72	32	C1R	1,901,221	3,837	322,378	0.7, 0.9 no	1.2, 0.8 no	1.2, 0.7 no
49	PKS 12564	81	None	C2L	627,318	2,487	None	0.7, 0 no*	0, 0 no	0.7, 0 no*
50	PKS 12671	81	311	C2L	927,088	8,259	1,067,728	0, 0 no	0, 0 no	0, 0 no
51	PKS 12709	81	311	C2L	1,031,825	7,872	1,189,154	0, 0 no	0, 0 no	0, 0 no
52	PKS 12806	81	311	C2L	1,314,426	8,574	1,475,006	0, 0 no	0, 0 no	0, 0 no
12	NRPS 2872	73	35	C2R	1,592,347	3,036	2,557,918	0.8, 0 no*	0, 0 no	0, 0.8 no
11	NRPS 2302	73	29	C2R	34,356	3,066	1,177,535	0, 0.4 no	0, 0.4 no	0, 0 no
36	PKS 10425	75	None	C3L	24,095	2,355	None
37	NRPS 10519	75	123	C3L	272,459	3,327	324,910	14, 0.1 yes	0.6, 0.1 no*	0.6, 14; +23 yes
38	PKS 10545	75	None	C3L	336,845	8,232	None	0.5, 0.3 no	1.0, 0.3 no*	1.0, 0.5 no*
39	PKS 10855	75	129	C3L	1,144,762	5,446	1,471,691	0.1, 0 no	0, 0 no	0, 0.1 no
40	PKS 284	75	None	C3L	2,307,795	7,005	None	18, 19 no	130, 19 yes	130, 19; −6.8 yes
53	NRPS 13549	78	77	C3R	1,209,037	3,216	1,030,927	0, 0 no	0, 0 no	0, 0 no
54	OmtA 13,921	78	77	C3R	2,199,609	1,634	15,777	300, 0 yes	78, 0 yes	78, 300 no*
55	PKS/NRPS	78	None	C3R	2,265,957	9,017	None	0.13, 0 no	17, 0 yes	17, 0 yes
54	PksA 13941	78	76	C3R	2,239,960	6,624	135,232	0, 0 no	0, 0 no	0, 0 no
10	Arp1 1614	74	222	C4L	901,342	462	959,335	70, 132 yes	0, 133 yes	0, 70; +>100 yes
5	PksP 617	82	44	C4R	1,664,602	6,651	228,115	0, 0 no	0, 0 no	0, 0 no
4	NRPS 544	82	None	C4R	1,442,264	7,962	None	0, 0 no	0.2, 2.2 no*	0.2, 0 no
2	DMAT 4300	82	44	C4R	1,066,674	1,340	833,092	25, 40 no	52, 40 no	52, 26 no*
3	NRPS 445	82	44	C4R	1,115,175	16,528	780,095	0.1, 0 no	1.5, 0.1 no*	1.5, 0.1 no*
1	PKS 290	82	44	C4R	639,511	7,320	1,256,446	2.9, 1.1 no*	2.2, 1.1 no*	2.2, 2.9 no
20	PKS 6282	77	25	C5L	2,153,493	7,923	854,574	0, 0 no	0, 0 no	0, 0 no
19	DMAT 6068	77	25	C5L	1,624,541	1,020	238,387	0.2, 0 no	0, 0 no	0, 0.2 no
30	NRPS 9020	80	139	C5R	302,181	5,256	1,501,541	0, 0 no	0, 0 no	0, 0 no
31	NRPS 9504	80	139	C5R	1,615,340	1,453	160,038	0, 0 no	0, 0 no	0, 0 no
32	C6TF 9637	80	253	C5R	1,928,138	2,531	126,829	0.2, 0.0.1 no*	0.4, 0.1 no*	0.4, 0.2 no*
33	PKS 9677	80	None	C5R	2,022,113	3,034	None
34	NRPS 10034	84	311	C6L	862,490	3,678	717,616	0, 0 no	0, 0 no	0, 0 no
35	NRPS 10170	84	215	C6L	1,198,962	3,129	115,582	0.1, 0.2 no*	0.6, 0.2 no*	0.6, 0.1 no*
25	ACV syn. 7086	79	46	C6R	1,819,006	11,325	173,928	0.1, 0.2 no*	0.3, 0.2 no	0.3, 0.1 no*
24	NRPS 6933	79	30	C6R	1,362,826	15,949	343,560	2.7, 1.0 no*	5.5, 1.0 no*	5.5, 2.7 no*
23	PKS/NRPS 6684	79	30	C6R	704,693	11,841	984,638	0,0 no	0.3, 0 no	0.3, 0 no
22	NRPS 6672	79	None	C6R	655,460	16,343	None
21	Gliotoxin 6442	79	30	C6R	77,944	461	1,686,932	2.6, 0 no*	1.6, 0 no*	1.6, 2.6 no
13	NRPS 3860	76	310	C7L	151,540	9,107	171,936	0, 0.1 no	0, 0.1 no	0, 0 no
14	IroE 4105	76	310	C7L	763,148	1,046	805,511	0, 0.5 no	0, 0.5 no	0, 0 no
15	DMAT 4549	76	310	C7L	1,948,240	1,375	2,001,813	0, 0 no	0, 0 no	0, 0 no
16	SidA 4719	76	310	C7L	2,481,875	1,575	2,561,114	22, 83 yes	21, 83 yes	21, 22 no
41	PKS 11482	87	208	C7R	153,643	5,571	123,539	0, 1 no*	0.7, 0.9 no	0.7, 0 no
6	NRPS 877	86	11	C8L	33,887	15,692	147,010	0, 0 no	0, 0 no	0, 0 no
7	PKS 914	86	11	C8L	459,100	1,463	44,826	0.2, 0 no	0.8, 0.1 no*	0.8, 0.2 no*
8	PKS/NRPS 1000	86	309	C8L	687,862	5,998	178,580	0.8, 0.5 no	0.1, 0.5 no*	0.1, 0.8 no*
9	NRPS 1058	86	309	C8L	815,096	23,346	313,447	0.1, 0.1 no	0.2, 0.2 no	0.2, 0.1 no*
48	NRPS 12152	85	None	C8R	1,479,298	3,093	None
47	NRPS 11911	85	69	C8R	896,447	3,314	1,018,151	0, 0 no	0, 0 no	0, 0 no
46	PKS 11896	85	69	C8R	849,586	9,017	919,461	0.008, 0 no	0.02, 0 no	0.02/0.008 no
45	NRPS 11844	85	69	C8R	710,225	3,304	774,162	6, 0 yes	1.1, 0 no*	1.1, 6.0; +5 yes
44	PKS 11689	85	69	C8R	296,613	7,791	328,410	0.2, 0.1 no*	0.1, 0.1 no	0.2, 0.2 no*
43	PKS 11650	85	None	C8R	188,716	1,505	None	0, 0 no	0, 0 no	0, 0 no
42	PKS 11622	85	69	C8R	121,844	6,466	153,847	0, 0 no	0, 0 no	0, 0 no

^aA single key gene in each of 55 secondary metabolite gene clusters identified previously in the A. flavus 3357 genome sequence (online) was used as a search query to analyze the A. parasiticus SU-1 scaffold sequences. The two species studied were A. flavus 3357 and A. parasiticus SU-1.

^bThe secondary metabolite (SM) cluster column shows the A. flavus gene cluster numbers.

^cThe AFLA gene no. column shows the predicted gene function and A. flavus gene designation for the key cluster gene. Abbreviations: NRPS, nonribosomal peptide; PKS, polyketide synthetase; DMAT, dimethylallyl tryptophan synthetase; C6TF, zinc binuclear cluster transcription factor; ACV syn., alpha aminoadipylcysteinyl valine synthase.

^dThe chromosome number and arm (left [L] or right [R]) in A. flavus 3357 and A. parasiticus SU-1 is shown.

^eThe key gene in 11 of 3,357 secondary metabolite clusters was not identified in A. parasiticus SU-1 and is shown as None.

^fRNA Seq analysis was conducted on duplicate samples of A. parasiticus SU-1 and A. flavus 3357 grown for 40 h in YES medium and on a single isolate of SU-1 grown on YEP medium. The level of expression of each of the key genes is shown in the three columns. A. parasiticus SU-1 grown on YES medium compared to A. parasiticus SU-1 grown on YEP medium (SU1 YES/YEP), A. flavus 3357 grown on YES medium compared to A. parasiticus SU-1 grown on YEP medium (3357 YES/SU-1 YEP), and A. flavus 3357 grown on YES medium compared to A. parasiticus SU-1 grown on YES medium (SU-1/3357 YES) are shown. The first value represents the expression level of the first member of the comparison pair, and the second number represents the expression level of the second member of the comparison pair. If the difference in expression between the first and second member of the pair is statistically significant, “yes” follows the values (based on the q value). If the difference in expression is not statistically significant, “no” follows the values. In some cases, the differences in expression were at least 2-fold, but this was not statistically significant; in this case, “no*” follows the values. +, up-regulation; −, down-regulation.

In contrast to manual analysis, antiSMASH software (default parameters) detected 93 secondary metabolite clusters in A. parasiticus SU-1, including 30 polyketide synthetase (PKS) clusters, 22 nonribosomal peptide (NRPS) clusters, 14 terpene clusters, 2 siderophore clusters, and 25 clusters designated “putative” or “other.” antiSMASH detected 87 secondary metabolite clusters in A. flavus 3357, including 28 PKS clusters, 22 NRPS clusters, 22 terpene clusters, 1 siderophore cluster, and 14 clusters designated “other.” We recently expanded this analysis to perform direct sequence comparison at the individual cluster level (J. Wee and J. E. Linz, unpublished data). Bioinformatics analysis support our initial observation that the proportions of the genome devoted to secondary metabolism are highly similar in A. parasiticus SU-1 and A. flavus 3357.

Comparison of A. parasiticus SU-1 and A. flavus 3357 gene expression under aflatoxin-inducing conditions.

The level of differential expression during growth on YES medium (40 h) was calculated for all genes in the A. parasiticus SU-1 genome (see Table S4 in the supplemental material). Of 13,290 total transcripts expressed by A. parasiticus SU-1, 1,408 (11%) exhibited significantly different levels of gene expression between A. parasiticus SU-1 and A. flavus 3357 during growth in YES medium at 40 h. This medium induces aflatoxin gene expression in both species, and aflatoxin gene expression increases at maximum rates between 30 and 40 h in YES medium (14). The number of genes exhibiting statistically significant differences in expression are much lower between two biological replicates of A. flavus 3357 (63 genes) or two biological replicates of A. parasiticus SU-1 (34 genes) analyzed in this study (see Table S5 in the supplemental material), suggesting that the observed differences in expression between A. parasiticus SU-1 and A. flavus 3357 are not due to variability in gene expression between biological replicates.

Six aflatoxin genes (cypA, aflT, nor-1, estA, hypC, and ordA) and one additional secondary metabolism gene (aroM) are expressed at significantly higher levels (YES medium, at 40 h) in A. parasiticus SU-1 than in A. flavus 3357 (Table 3). In support of this observation, 13 additional genes in the aflatoxin gene cluster (cluster 54) are expressed at a minimum of 2-fold-higher levels in A. parasiticus SU-1 than in A. flavus 3357 including fas-2, a conserved hypothetical protein, omtB, avfA, hypB, verB, avnA, hypE, ordB, omtA, vbs, cypX, and moxY (Table 3 and Fig. 3); these differences in expression were not statistically significant. In contrast, the gene encoding the hybrid polyketide synthase/nonribosomal peptide synthase in the CPA gene cluster 55 and three stress response genes (Hsp70, MnSOD, and xylulose reductase) are expressed at significantly higher levels in A. flavus 3357 than in A. parasiticus SU-1. These data may help explain differences in the quantity of aflatoxin synthesized by the two species under our laboratory conditions (Fig. 4) and differences in CPA produced by A. flavus 3357. In support of these data, we did not detect the CPA NRPS/PKS (gene designation AFLA_13949) in the A. parasiticus de novo genome sequence. We also did not detect transcripts for adhA and, aflS (in cluster 54) as well as major facilitator superfamily (MFS) multidrug transporter, flavin adenine dinucleotide (FAD)-dependent oxidoreductase, and dimethylallyl tryptophan synthase (in cluster 55, the CPA cluster) in A. parasiticus SU-1.

TABLE 3.

Expression of selected genes involved in secondary metabolism, stress response, and development

Gene group	Gene or protein name(s)	AFLA gene designation	Ratio of expressiona		Level of expression (SU-1/3357 YES)b
			SU1 YES/YEP	3357 YES/SU-1 YEP
Aflatoxin	aflF, norB	13944	+>100*
	aflU, cypA	13943	+528	+89	2,600, 357; +7.3
	aflT	13942	+528	+89	2,600, 357; +7.3
	hypC	13940	+528	+89	2,600, 357; +7.3
	aflD, nor-1	13939	+528	+89	2,600, 357; +7.3
	aflA, fas-2	13938	+10	+8
	aflB, fas-1	13937	+>10*
	aflR	13936	+1.8	+3
	aflJ, estA	13932	+1,671	+960	664, 288; +2.3
	hypE	13929	+7,361	+960
	hypD	13927		+70
	aflG, avnA	13926	+7,361	+960
	aflL, verB	13925	+7,361	+960
	hypB	13924	+7,361	+960
	aflI, avfA	13923	+7,361	+960
	aflO, omtB	13922	+7,361	+960
	aflP, omtA	13921	+>1,000	+>1,000
	aflQ, ordA	13920	+>1,000	+>1,000	106, 22; +4.8
	aflK, vbs	13919	+346	+100
	aflV, cypX	13918	+>1,000	+>1,000*
	aflW, moxY	13917	+346	+100
	aflX, ordB	13916	+>1,000*	+>1,000*
	aflY, hypA	13915	+>1,000*	+>1,000*
	nadA	13914	+346	+100
	hxtA	13913	−10*	+>100*
	glcA	13912	+346	+100
	sugR	13911	−10*
Stress response	Mn SOD	3342	+129	+700	167, 775; −4.6
	Hsp70	849			11, 24; −2.2
	Xylulose reductase	1558		+2.5	132, 282; −2.1
Conidiation	Medusa	13641		+>100*
	RosA	2193	+>100*
	AbaA	2962		+3*
	BrlA	8285	+2*	+>100*
Secondary metabolism	AroM	1768	+2		16, 7; +2.3
	PKS_NRPS (CPA)	13949		+>100	0.1, 17; −170
C6 transcription factors	AflR	13936		+3
	Unknown	6437	−3*	−>10*
	Unknown	4737	−2*
	Cluster 29	8409	+>10*	+>100
	Cluster 11	2304	−5	−10
Other transcription factors	AtfB	9401		+4
Housekeeping genes	Citrate synthase	4929	−25	−40

^aThe ratios of raw expression data of A. parasiticus SU-1 grown for 40 h in YES medium to strain SU-1 grown for 40 h in YEP medium (SU-1 YES/YEP) and A. flavus 3357 grown for 40 h in YES medium to A. parasiticus SU-1 grown for 40 h in YEP medium (3357 YES/SU-1 YEP) are shown. All values reported in these two columns are significant (q < 0.05) except for the values with asterisks. The values with asterisks display trends in differential expression, but they are not statistically significant (q > 0.05). Symbols; +, upregulated in YES or YEP medium; −, downregulated in YES or YEP medium.

^bThe level of expression of A. parasiticus SU-1 grown for 40 h in YES medium and the level of expression of A. flavus 3357 grown for 40 h in YES medium (SU-1/3357 YES) are shown. The first value represents the raw expression data for strain SU-1 grown in YES medium at 40 h, and the second value represents the raw expression data for strain 3357 grown in YES medium for 40 h. The ratio of these values determines the fold upregulation (+) or downregulation (−), which is shown after the first two values. All values reported in this column are statistically significant.

FIG 3.

RNA Seq analysis of gene expression in secondary metabolite clusters 54 and 55 in A. parasiticus SU-1 and A. flavus 3357. (A) Differential expression of all genes during growth in YES medium for 40 h. The relative level of differential expression (log base 2) between A. parasiticus SU-1 relative to A. flavus 3357 is indicated by the distance above (upregulation) and below (downregulation) the zero difference horizontal line. The number of transcripts is shown on the x axis. (B) Differential expression of specific genes in secondary metabolism clusters 54 and 55. Genes that are upregulated (blue diamonds) and genes that are downregulated (red diamonds) are indicated. The common gene name(s) is presented next to the colored diamond and is based on annotations in the A. flavus genome browser. In specific cases, RNA Seq analysis presented aggregate expression data for closely spaced genes within cluster 54 (e.g., omtB, avfA, hypB, verB, avnA, and hypE [omtB/avfA/hypB/verB/avnA/hypE]). Genes included within these aggregates are closely spaced and encoded on the same strand of DNA. Presumably RNA SEQ was unable to resolve independent data for these groups of genes. (C) Bar graph presentation of differential expression data in panel B.

FIG 4.

Aflatoxin accumulation in A. parasiticus SU-1 and A. flavus 3357 grown in culture. A. parasiticus SU-1 and A. flavus 3357 were grown in triplicate under standard aflatoxin-inducing (YES medium, 40 h) and aflatoxin-noninducing (YEP medium, 40 h) culture conditions as described in Materials and Methods. Total aflatoxin was extracted from cultures and analyzed by thin-layer chromatography as described in Materials and Methods. Lane S illustrates migration of aflatoxin standards B₁, B₂, G₁, and G₂. An unidentified metabolite appears in all cultures above the aflatoxin spots.

Of the 44 secondary metabolite gene clusters detected manually in A. parasiticus, 26 were not expressed in either A. flavus 3357 or A. parasiticus SU-1 under the aflatoxin-inducing or -noninducing conditions utilized in this study (Table 2), suggesting that these gene clusters are either not expressed at significant levels or are expressed under growth conditions that these fungi encounter in the wild. The NRPS in clusters 37 and 45 as well as Arp1 in cluster 10 were expressed at significantly higher levels after 40 h of growth of A. parasiticus SU-1 in YES medium, while the PKS in cluster 40 was expressed at significantly higher levels in A. flavus 3357 under these same conditions (Fig. 3 and Table 3). We also observed at least 2-fold differences in expression levels between A. parasiticus SU-1 and A. flavus 3357 in 13 additional secondary metabolism gene clusters (Table 3), providing evidence that these species differ greatly in the type and quantity of secondary metabolites synthesized under the specified growth conditions.

Gene expression in A. parasiticus SU-1 and A. flavus 3357 was also compared during growth in YES medium (aflatoxin-inducing growth conditions) and YEP medium (aflatoxin-noninducing growth conditions) (see Table S6 in the supplemental material) with a focus on aflatoxin, secondary metabolism, and cellular defense/stress response genes (Table 3). A total of 1,284 A. parasiticus SU-1 genes and 1,802 A. flavus 3357 genes exhibited significant differences in gene expression when grown in YES medium or in YEP medium at 40 h (see Table S6 in the supplemental material). We analyzed only a single biological replicate of strain SU-1 on YEP medium for these comparisons, so these data are preliminary in nature. However, most genes within the aflatoxin gene cluster are upregulated in YES medium as expected for both A. flavus 3357 and A. parasiticus SU-1 (Fig. 3 and Table 3), so it is likely that these trends in expression will be confirmed upon further analysis.

Identification of “uniquely expressed genes” in A. parasiticus.

We observed that 860 A. parasiticus SU-1 loci have no homolog in the A. flavus 3357 genome sequence based on direct DNA sequence comparison between the genomes and RNA Seq analysis of gene expression on YES medium and/or YEP medium (see Tables S4 and S6 in the supplemental material). The absence of these loci in the A. flavus 3357 genome may imply that these loci are unique A. parasiticus SU-1 gene sequences. Alternatively, these loci may be located within gaps in the A. flavus 3357 genome sequence. Of these 860 loci, the difference in the level of expression between A. flavus 3357 and A. parasiticus SU-1 for 318 of these loci is statistically significant. For 213 of these loci, RNA transcripts (expression) could not be detected in A. flavus 3357, while transcripts were detected at significant levels in A. parasiticus SU-1 (Fig. 5 and Table 4). We identified gene function for 68 of these genes by BLAST analysis of the translated protein products and then assigned predicted function by KOBAS GO analysis. Of particular importance to the current study, we identified 9 fungus-specific transcription factors, 4 enzymes involved in secondary metabolism, and 14 enzymes involved in stress response/cellular defense that are unique to A. parasiticus SU-1 and likely contribute to phenotypic and biochemical differences observed in the two species. Surprisingly, no known function was assigned to 145 of the 213 uniquely expressed genes (68%) based on BLAST analysis of the NCBI database (Fig. 5), suggesting that they may contribute unique functions in the biology of A. parasiticus SU-1.

FIG 5.

KOBAS GO pathway analysis of uniquely expressed genes in A. parasiticus SU-1. Of 213 uniquely expressed A. parasiticus SU-1 genes, we were unable to assign predicted functions for 145 genes (68%) based on BLAST analysis of the NCBI genome database. The remaining 68 genes were assigned predicted functions and assigned to functional groups. Functional group annotations appear on the y axis as follows: Unassigned, transcripts with no known or assigned function; Other, transcripts with predicted functions other than SR, TF, and SM; SM, secondary metabolism; TF, transcription factors; SR, oxidative stress response. Each bar represents the number of genes in each functional group.

TABLE 4.

Predicted function of selected proteins by functional category

Gene_IDa	Protein function	Categoryb
XLOC_003627	3-Hydroxybenzoate 6-hydroxylase 1	SR
XLOC_001416	Acyl-oxidase	SR
XLOC_010908	Acyl-oxidase	SR
XLOC_011069	Acyl-oxidase	SR
XLOC_006050	Aldo keto reductase	SR
XLOC_004307	Cytochrome p450	SR
XLOC_013229	Cytochrome p450 alkane	SR
XLOC_013253	Glyoxalase bleomycin resistance protein dioxygenase	SR
XLOC_001771	Phytanoyl-dioxygenase family protein	SR
XLOC_011364	Salicylate hydroxylase	SR
XLOC_003035	Short-chain dehydrogenase	SR
XLOC_008563	Short-chain dehydrogenase reductase family	SR
XLOC_009101	Zinc-type alcohol dehydrogenase-like protein	SR
XLOC_000430	Dimeric dihydrodiol dehydrogenase	SR
XLOC_001156	C6 finger domain-containing protein	TF
XLOC_011187	C6 transcription factor	TF
XLOC_008080	C6 transcription factor	TF
XLOC_010864	C6 transcription factor	TF
XLOC_000259	C6 zinc finger domain-containing protein	TF
XLOC_000432	Fungus-specific transcription factor	TF
XLOC_013226	Fungus-specific transcription factor	TF
XLOC_006054	Fungus-specific transcription factor domain-containing protein	TF
XLOC_010866	Zn-C6 fungal-type DNA-binding domain	TF
XLOC_004365	ent-kaurene synthase	SM
XLOC_013227	NRPS-like enzyme	SM
XLOC_003657	O-methyltransferase involved in polyketide biosynthesis	SM
XLOC_008561	Polyketide synthase	SM

^aXLOC, A. parasiticus SU-1 gene designation.

^bSR, stress response; TF, transcription factor; SM, secondary metabolism.

DISCUSSION

The focus of this ongoing work was to identify the genetic basis for the observed species differences between A. parasiticus SU-1 and A. flavus 3357. Overall, the number of genes identified in our analysis of the A. parasiticus SU-1 genome (13,290) was similar to that identified in A. flavus 3357 (13,487). Even though their genome sequences and chromosome structure are similar, we observed statistically significant levels of differential expression between A. flavus 3357 and A. parasiticus SU-1 during growth in YES medium for 40 h and during growth on YES medium or YEP medium for 40 h. Furthermore, expression of 2,930 A. parasiticus SU-1 genes could not be detected in YES medium, and expression of 3,836 genes could not be detected on YEP growth medium. Since we analyzed only a single strain of each species, the results we present provide a starting point for a broader analysis of many different strains of each species to determine whether differences we observed in the current work extend beyond the two strains we analyzed.

One short-term goal of the current work was to explain specific species differences in mycotoxin biosynthesis, including the quantity and type of aflatoxin produced in culture as well as the synthesis of cyclopiazonic acid (CPA). We observed significantly higher levels of expression of 6 key genes (cypA, aflT, nor-1, estA, hypC, and ordA) in A. parasiticus SU-1 as well as a minimum 2-fold increase in expression of 13 other aflatoxin genes that could in part explain species differences in aflatoxin biosynthesis, and future work can focus on expanding the focus of this analysis to additional strains of A. flavus and A. parasiticus. Upregulation of all 19 of these aflatoxin genes in A. parasiticus SU-1 could account for higher levels of aflatoxin synthesized by this species, while higher levels of ordA and cypA specifically (which mediate conversion of B aflatoxins to the corresponding G aflatoxins) could explain why most strains of A. parasiticus synthesize both B and G aflatoxins, while most strains of A. flavus synthesize predominantly B aflatoxins. We also detected significantly higher expression of the hybrid PKS/NRPS in A. flavus 3357 that catalyzes the initial steps in the synthesis of CPA in many A. flavus strains. The CPA PKS/NRPS was expressed at up to 100-fold-higher levels in A. flavus 3357 than in A. parasiticus SU-1. In support of this observation, this hybrid gene and three other genes in cluster 55 were not detected in the assembled A. parasiticus SU-1 genome.

We also compared the presence and location of 55 secondary metabolism clusters present in the A. flavus 3357 genome. While 42 of these clusters were detected in both organisms at the same approximate location, 1 cluster was present on a different chromosome, perhaps providing evidence for intragenomic recombination leading to translocation of this cluster. During analysis of the 318 loci that were present in the A. parasiticus SU-1 genome but not A. flavus 3357, we detected additional enzymes like PKS and NRPS, so it appears that both species are able to synthesize unique secondary metabolites. Further analysis of the A. flavus 3357 genome must be conducted to determine whether these genes were present but detected as part of the original sequence analysis or if they are indeed unique to A. parasiticus SU-1.

antiSMASH is a simple and rapid bioinformatics analysis for identification and functional annotation for secondary metabolite gene clusters in bacterial and fungal genome sequences (10). antiSMASH algorithms are based on specific hallmarks (functional motifs) and sequence identity in clusters of secondary metabolism Clusters of Orthologous Groups (smCOGs). This software program was reported to have 97% reliability when it was used to compare 484 cloned fungal and bacterial secondary metabolism gene clusters in the GenBank database. While we manually detected 44 clusters based on a single key enzyme present in a specific pathway, antiSMASH expanded upon this analysis and detected 93 clusters based on a collection of enzymes present in the same cluster. antiSMASH detected similar numbers of PKS and NRPS clusters in A. parasiticus SU-1 (30 PKS and 22 NRPS) and A. flavus 3357 (28 PKS and 22 NRPS), and approximately 60% of the secondary metabolite clusters are committed to synthesis of PKS and NRPS in both species. Of particular interest, there was a clear difference in the number of terpene clusters in A. parasiticus SU-1 (14 clusters) and A. flavus 3357 (22 clusters). Analyzing how the observed differences in the number and predicted function of terpene and siderophore clusters impact species phenotype will be one focus of future research.

Finally, we detected 213 genes that appear to be “uniquely expressed” in A. parasiticus SU-1. These genes and their transcripts were not identified in A. flavus 3357 as part of the current study or in the A. flavus 3357 genome available online. KOBAS GO Pathway analysis enabled us to place 27 of these 213 genes into functional groups, including secondary metabolism (4 genes), cellular defense/stress response (14 genes), and C6 (zinc binuclear cluster) transcription factors (9 genes) (15, 16).

Members of the C6 transcription factor family are particularly interesting, because they are specific to fungi and are strongly associated with many secondary metabolism gene clusters in both A. flavus and A. parasiticus. In the current study, we manually identified 13 different C6 transcription factors associated with secondary metabolite gene clusters in both A. flavus 3357 and A. parasiticus SU-1. AflR, one of those 13 C6 transcription factors, is a key positive regulator of expression of most of the aflatoxin genes (17); aflR is clustered with the other aflatoxin genes on chromosome 3 in A. parasiticus SU-1 and A. flavus 3357. In parallel studies, we used chromatin immunoprecipitation-sequencing (ChIP-Seq) analysis to demonstrate that the basic leucine zipper (bZIP) transcription factor AtfB interacts with promoters of several different C6 transcription factors in A. flavus 3357, suggesting that AtfB plays a direct role in regulation of secondary metabolism via activation of expression of C6 transcription factors (Wee et al., unpublished). We previously demonstrated that AtfB and AflR physically interact, suggesting that AtfB may help recruit the C6 transcription factors to promoters of secondary metabolism genes. Yin et al. demonstrated that another bZIP designated RsmA, binds to two sites in the AflR promoter, helping to activate its expression in Aspergillus nidulans (18). These observations suggest that the C6 transcription factors play an important role in regulating secondary metabolism and that we could target them to modulate secondary metabolite expression without affecting the host plant or humans that consume crops associated with mycotoxins.

The 27 “uniquely expressed” genes and their functional networks described briefly above will be the focus of future efforts in analysis of the genetic basis for biochemical and morphological differences between these closely related species.