Dual Transcriptome of Post-Germinating Mutant Lines of Arabidopsis thaliana Infected by Alternaria brassicicola

Abstract

Alternaria brassicicola is a seed-borne pathogen that causes black spot disease in Brassica crops, yet the seed defense mechanisms against this fungus remain poorly understood. Building upon recent reports that highlighted the involvement of indole pathways in seeds infected by Alternaria, this study provides transcriptomic resources to further elucidate the role of these metabolic pathways during the interaction between seeds and fungal pathogens. Using RNA sequencing, we examined the gene expression of glucosinolate-deficient mutant lines (cyp79B2/cyp79B3 and qko) and a camalexin-deficient line (pad3), generating a dataset from 14 samples. These samples were inoculated with Alternaria or water, and collected at 3, 6, and 10 days after sowing to extract total RNA. Sequencing was performed using DNBseq™ technology, followed by bioinformatics analyses with tools such as FastQC (version 0.11.9), multiQC (version 1.13), Venny (version 2.0), Salmon software (version 0.14.1), and R packages DESeq2 (version 1.36.0), ClusterProfiler (version 4.12.6) and ggplot2 (version 3.4.0). By providing this valuable dataset, we aim to contribute to a deeper understanding of seed defense mechanisms against Alternaria, leveraging RNA-seq for various analyses, including differential gene expression and co-expression correlation. This work serves as a foundation for a more comprehensive grasp of the interactions during seed infection and highlights potential targets for enhancing crop protection and management.

Dataset: The RNA-seq data were submitted to the NCBI. This dataset is publicly available through the GEO database with accession number GSE214602 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE214602, accessed on 25 August 2024) and all the supplementary tables can be accessed via Mendeley data (https://data.mendeley.com/datasets/c9d453jtwr/2, accessed on 25 August 2024).

Dataset License: license under which the dataset is made available (CC0, CC-BY, CC-BY-SA, CC-BY-NC, etc.).

1. Summary

Secondary metabolites such as indole glucosinolates and camalexin [1] both originate from tryptophan (Trp) and play a crucial role in the defense mechanisms of Brassicaceae plants against certain phytopathogens [1,2,3,4]. One such seed-borne pathogen is Alternaria brassicicola (referred to as Alternaria hereinafter), a necrotrophic fungus responsible for inducing black spot disease in Brassica crops [5,6]. In plants, the antifungal activity from these metabolites has been extensively documented [1,3,7]; however, their role in seed defense remains poorly studied.

Recently, our research group has been studying the seed defense mechanisms against fungal pathogens using the Arabidopsis thaliana/Alternaria brassicicola pathosystem [6,8]. We focused on developmental stages closely related to seed germination, specifically germinating seeds, seedling emergence (stages just before and after radicle protrusion through seed covering layers), and establishment (autotrophic stage), corresponding to 3, 6, and 10 days after sowing (DAS), respectively [8].

First, a transcriptomics experiment was conducted on Arabidopsis Col0 genotype (wild-type—WT) infected by Alternaria [9]; subsequently, phenotyping analyses were performed to assess colonization rates and necrosis development in defense-defective mutant seeds (WT was used as the control) [8]. These studies enabled the identification of a specific immune response in seeds against Alternaria. We found that the necrotrophic fungus might modulate the seed transcriptome to facilitate colonization and induce necrosis development through the glucosinolate pathway. Additionally, the camalexin-deficient line displayed a necrosis pattern like the WT strain [8]. To enhance our comprehension of the involvement of glucosinolate pathways in the seed immune response, we investigated the transcriptomic profiles of the following mutant Arabidopsis seeds: a cyp79B2 cyp79B3 double mutant, which is defective in indole glucosinolates (cyp79B2/B3) [10]; a cyp79B2 cyp79B3 myb28 myb29 quadruple mutant, which is defective in both indole and aliphatic glucosinolates (qko) [11]; and a camalexin-deficient mutant (pad3) [12]. Seeds were subjected to infection with an Alternaria inoculum (10⁴ conidias/mL) or imbibed in water as a control. At 3, 6, and 10 DAS, RNA isolation was conducted, and samples with optimal quality were sent to the Beijing Genomics Institute for sequencing using DNBseq™ technology. Our dataset contains 84 clean reads from 14 samples. A sequenced reads quality filter was performed using FastQC version 0.11.9 [13] and MultiQC version 1.13 [14] was performed in all files. High-quality reads were mapped on the Arabidopsis [15] and the Alternaria [16] reference genomes using Salmon version 0.14.1 [17]. The expression change was calculated by DESeq2 packages version 1.36.0 [18]. Gene expression changes were inferred by comparing infected and healthy WT or mutant seed expression values, as well as performing pairwise comparisons of expression values between mutant and WT genotypes, both under infected conditions. Functional annotation was obtained using the package ClusterProfiler version 4.12 [19]. This study provides RNA-seq data as resources that can significantly contribute to elucidating the seed defense mechanism against fungal necrotrophs. By utilizing glucosinolate mutants, these expression data enable a deeper exploration of the role of the indole glucosinolate pathway in this intricate defense mechanism.

2. Data Description

2.1. Identification of Dual Transcriptome

After sequencing, the reads were filtered, and high-quality reads were aligned and mapped to the Arabidopsis [15] and Alternaria [16] reference genomes. The completed RNA-seq data were deposited in the NCBI Sequence Read Archive (SRA) database under the repository name NCBI GEO with the data identification number GSE214602 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE214602, accessed on 25 August 2024).

Table 1. Summary of mapped reads to reference genomes of both Arabidopsis and Alternaria. Mutant lines: cyp79B2/B3, a double mutant defective in indole glucosinolate [10]; qko, a quadruple mutant defective in indole and aliphatic glucosinolates [11]; and pad3, a camalexin-deficient mutant [12]. Inoc: seed inoculated with Abra43; water: seed without fungal inoculum; developmental stages: 3, 6, and 10 days after sowing (DAS); REP: biological replicate. % Aligned: % mapped reads; M Aligned: mapped reads (millions); M Seqs: total sequences (millions).

Sample Name	M Seqs	Mapped Reads Percentages to Arabidopsis Genome		Mapped Reads Percentages to Alternaria Genome
		% Aligned	M Aligned	% Aligned	M Aligned
cyp_inoc_3d_REP1	25.2	95.1%	24.0	1.0%	0.3
cyp_inoc_3d_REP2	50.9	85.5%	43.5	8.2%	4.2
cyp_inoc_3d_REP3	51.2	89.8%	46.0	4.9%	2.5
cyp_inoc_6d_REP1	51.0	72.7%	37.1	17.4%	8.9
cyp_inoc_6d_REP2	52.5	44.7%	23.5	37.6%	19.7
cyp_inoc_6d_REP3	50.2	62.8%	31.5	24.5%	12.3
cyp_inoc_10d_REP1	50.8	46.4%	23.6	35.8%	18.2
cyp_inoc_10d_REP2	51.8	45.5%	23.6	36.2%	18.8
cyp_inoc_10d_REP3	51.9	27.2%	14.1	48.4%	25.1
qko_inoc_3d_REP1	52.1	30.7%	16.0	47.5%	24.8
qko_inoc_3d_REP2	51.9	38.5%	20.0	41.2%	21.4
qko_inoc_3d_REP3	51.9	24.4%	12.7	52.2%	27.1
qko_inoc_6d_REP1	52.1	16.9%	8.8	55.6%	29.0
qko_inoc_6d_REP2	52.3	28.4%	14.9	48.0%	25.1
qko_inoc_6d_REP3	52.1	9.1%	4.7	58.7%	30.6
pad3_inoc_3d_REP1	50.9	65.6%	33.3	20.3%	10.3
pad3_inoc_3d_REP2	50.7	34.1%	17.3	40.4%	20.5
pad3_inoc_3d_REP3	51.0	76.0%	38.7	13.3%	6.8
pad3_inoc_6d_REP1	50.4	83.3%	41.9	9.0%	4.5
pad3_inoc_6d_REP2	50.1	41.1%	20.6	38.3%	19.2
pad3_inoc_6d_REP3	51.3	65.6%	33.6	20.8%	10.7
cyp_water_3d_REP1	25.2	96.4%	24.3	0.0%	0.0
cyp_water_3d_REP2	25.2	95.4%	24.1	0.0%	0.0
cyp_water_3d_REP3	25.2	96.7%	24.4	0.0%	0.0
cyp_water_6d_REP1	25.2	95.4%	24.0	0.0%	0.0
cyp_water_6d_REP2	25.2	95.0%	23.9	0.0%	0.0
cyp_water_6d_REP3	25.2	96.1%	24.2	0.0%	0.0
cyp_water_10d_REP1	25.3	96.2%	24.3	0.0%	0.0
cyp_water_10d_REP2	25.2	96.4%	24.3	0.0%	0.0
cyp_water_10d_REP3	25.3	96.2%	24.3	0.0%	0.0
qko_water_3d_REP1	25.2	96.4%	24.3	0.0%	0.0
qko_water_3d_REP2	25.2	96.5%	24.3	0.0%	0.0
qko_water_3d_REP3	26.4	96.7%	25.6	0.0%	0.0
qko_water_6d_REP1	25.8	94.2%	24.3	0.0%	0.0
qko_water_6d_REP2	26.5	95.8%	25.4	0.0%	0.0
qko_water_6d_REP3	26.5	95.3%	25.2	0.0%	0.0
pad3_water_3d_REP1	26.0	94.8%	24.7	0.0%	0.0
pad3_water_3d_REP2	25.9	94.2%	24.4	0.0%	0.0
pad3_water_3d_REP3	26.0	95.7%	24.9	0.0%	0.0
pad3_water_6d_REP1	25.9	95.5%	24.7	0.0%	0.0
pad3_water_6d_REP2	26.0	95.8%	24.9	0.0%	0.0
pad3_water_6d_REP3	25.8	95.3%	24.6	0.0%	0.0

provides a summary of the mapped reads to their respective reference genomes for both the plant and the necrotrophic fungus, presenting the percentage of reads mapped for each biological replicate of each one. Notably, the samples from the inoculation group exhibit variations in mapped reads between Arabidopsis and Alternaria across different post-infection time points. Mapped reads to Arabidopsis remain relatively high, indicating successful mapping to the plant genome. In contrast, the lower percentage of reads mapped to the Alternaria genome may be due to the preliminary state and ongoing curation of the Alternaria Abra43 genome [16], which could have influenced mapping accuracy.

Additionally, control samples exposed to water consistently exhibit a high number of mapped reads for the Arabidopsis reference genome, confirming the absence of infection. Also, no reads were aligned to the Alternaria genome in the control samples, supporting the specificity of the alignment process and validating the absence of fungal contamination in the control conditions.

2.2. Differentially Expressed Genes from Infected and Healthy Arabidopsis Mutant Seeds at 3 and 6 DAS

After the quality control of reads and reference base assemblies, differentially expressed genes (DEGs) were obtained by comparing between infected and healthy mutant seeds at 3 and 6 DAS using the DESeq2 tool [18]. Genes with log2FC > 1 or  <−1 and Benjamani–Hochberg score < 0.05 were considered as DEGs. A total of 256, 2954, and 1942 DEGs were identified, respectively, for cyp79B2/B3, qko, and pad3 at 3 DAS and 5345, 3949, and 6768 at 6 DAS. The lists of these DEGs can be found in Table S1, including information related to gene’s foldchange, “Gene Symbol” and “Gene TAIR Computational Description” from The Arabidopsis information Resource—TAIR [20]—among others. The total number DEGs, classified as both up- and downregulated, for each mutant line at 3 and 6 DAS is shown in

Table 2. Total number of differentially expressed genes (DEGs) in cyp79B2/B3, qko, and pad3 Arabidopsis mutant lines. DEGs from comparison between infected and healthy mutant seeds (log2FC > 1 or  <−1 and Benjamani–Hochberg adjusted p-value < 0.05).

Mutant Lines	DEGs	3 DAS	6 DAS
cyp79B2/B3	Upregulated	237	3362
	Downregulated	19	1983
qko	Upregulated	1817	2092
	Downregulated	1137	1857
pad3	Upregulated	797	3703
	Downregulated	1145	3065

.

Previously, we analyzed sequences from WT seeds infected with Alternaria [9]. Sampling, sequencing, and bioinformatic procedures were close to those of the current study; we overlapped the DEGs identified in the WT with those in the Arabidopsis mutants cyp79B2/B3, qko, and pad3 using Venn diagrams [21]. The number of DEGs within the intersections, both up- and downregulated, varies based on the plant’s developmental stage and the presence of seed mutations. Notably, our observations revealed that the WT shares several common genes with the pad3 mutant at both 3 and 6 DAS (Figure 1).

The functional annotation of DEGs obtained by the comparison between healthy and infected seeds allowed us to compare the enrichment of GO terms in the WT genotype and in the different mutants cyp79B2/B3, qko, and pad3 (Figure 2). The observed differences highlight that the mutated genes participate in the response to Alternaria infection in a wild genetic background of the Col0 ecotype. This observation is in agreement with the differences in infection symptoms previously observed in young seedlings [8]. A previous transcriptomic study carried out on the WT Col0 genotype on rosette leaves infected with Alternaria identified 505 genes involved in the response to infection [22]. Among these genes, a number of 1, 382, and 400 genes were found to induced, respectively, in our samples at 3, 6, and 10 DAS [8]. Clearly, the response to Alternaria at the germination stage (3 DAS) differs radically from that observed after germination (6 DAS) in the established seedling (10 DAS). These last two stages present a response similar to the response observed in infected rosette leaves [8,22].

Moreover, Narusaka et al. [23] illustrate that the transcriptome of Alternaria-infected leaves in the pad3 mutant, compared to the WT Col0 genotype, shows a defect in the induction of genes in the salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) defense pathways. Both genotypes exhibit very similar hypersensitive response (HR) reactions at the infected leaf level; however, the pad3 mutant shows increased susceptibility to Alternaria colonization. Narusaka et al. [23] hypothesize that pad3 not only affects camalexin production but also delays the induction of Alternaria-responsive genes mediated by the HR reaction. Other studies at the post-germination stage of the seed (6 DAS) found no differences in symptoms between the WT and pad3 genotype [8]. The RNA-seq data described here (

Table 3. Defense response genes induced by Alternaria. A list of defense-related genes was constructed using defense response terms from Gene Ontology. These genes were then identified among the upregulated genes in the WT and pad3 at 3 and 6 DAS and compared between both genotypes. JA: jasmonic acid; ET: ethylene; SA: salicylic acid; ROS: reactive oxygen species; GSL: glucosinolates.

Defense Pathways and Metabolites	No. of Genes at 3DAS		No. of Genes at 6DAS
	WT	pad3	WT	pad3
JA	1	1	22	15
ET	41	22	118	104
SA	92	71	273	252
ROS	21	13	61	55
Indol	57	48	113	99
Phytoalexin	11	11	20	19
GSL	11	5	21	17

) indicate that the genes involved in the SA, JA, and ET defense pathways are indeed induced in greater numbers in the WT genotype compared to the pad3 mutant at both the pre-germination stage (3 DAS) and at the post-germinative stage. At the post-germinative stage (6 DAS), the number of induced defense genes increases and the difference in the number of induced genes between the WT and pad3 is also observed, but to a lesser extent. This suggests that, from the onset of development during germination, the pad3 mutation results in a reduced Alternaria-mediated induction of SA, JA, and ethylene defense pathways. At the stage of seedling emergence, the difference between the two genotypes becomes smaller.

2.3. Comparative Analysis of Gene Expression Changes in Alternaria-Infected Seeds of Arabidopsis Mutants and Wild-Type at 3 and 6 Days After Sowing

Gene expression changes for Arabidopsis (Table S2) and Alternaria (Table S3) are provided as separate datasets. All expression changes correspond to pairwise comparisons between the mutants and the WT Arabidopsis seeds infected by Alternaria. Figure 3 illustrates the number of genes categorized as downregulated, non-significant, and upregulated at 3 and 6 DAS.

2.4. Gene Expression Changes in cyp79B2/B3 Arabidopsis Mutant at 10 DAS in Both Plant and Fungi Transcriptomes

Gene expression changes at 10 DAS were only obtained for the plant mutant cyp79B2/B3 (Table S4), and these results are illustrated through volcano plots (Figure 4). Figure 4A,B depict genes from Arabidopsis while Figure 4C corresponds to genes from Alternaria, both in the mutant cyp79B2/B3. Gene expression changes for Figure 4A,C were obtained by comparing between infected and healthy cyp79B2/B3 seeds, while Figure 4B values were calculated through a pairwise comparison between cyp79B2/B3 and the WT-infected seeds.

3. Methods

3.1. Plant Material and Treatments

Seeds from three different mutant lines were used in this study: cyp79B2/B3, an indole glucosinolate mutant (cyp79B2 cyp79B3) [10], qko, an indole and aliphatic glucosinolate mutant (cyp79B2 cyp79B3 myb28 myb29) [11], and pad3, a camalexin-deficient mutant [12]. Mutant seeds were derived from the Columbia—Col0—genetic background, and the WT Arabidopsis accession Col-0 was used as the control [9]. All seeds were obtained from plants grown under controlled conditions (20 °C, 16 h light photoperiod, and 70% relative humidity) at the Institut de Recherche en Horticulture et Semences (IRHS), Beaucouzé, France [9]. In brief, the seeds were surface-sterilized and infected with an inoculum of Alternaria, strain Abra43, at a concentration of 10⁴ conidias/mL or imbibed in water as controls. All samples were incubated in a controlled growth chamber for 3, 6, and 10 days, following the protocol described by Ortega-Cuadros et al. [9]. Three biological replicates were collected for each condition and preserved at −80 °C until further use.

3.2. RNA Extraction and Sequencing

RNA isolation was performed for each sample using 20 mg of seeds. Samples from the cyp79B2/B3 mutants were collected at 3, 6, and 10 DAS, while samples from qko and pad3 mutants were collected only at 3 and 6 DAS. Total RNA was extracted using the NucleoSpin^® RNA Plus kit (Macherey-Nagel, Düren, Germany) with previous modifications [8]. The quantity and integrity (Ribosomal Integrity Number—RIN) for RNA were measured with a NanoDrop ND-100 (NanoDrop Technologies, DE, USA) and 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA), respectively. Samples with optimal parameters were sent to Beijing Genomics Institute—BGI (https://www.bgi.com, accessed on 25 August 2024)—Hong Kong, for cDNA library preparation, and paired-end sequencing (PE100, 40 M) using the DNBseq™ technology.

3.3. Data Analysis

After sequencing, BGI filtered the raw reads, removing adaptor sequences, contamination, and low-quality reads. Furthermore, the quality control of the raw reads was checked and summarized with the FastQC version 0.11.9 [13] and MultiQC tools version 1.13 [14]. Reads with Phred scores ≥ 35 were selected as high-quality reads. The quasi-mapping alignment method from Salmon version 0.14.1 [17] was employed for mapping the reads corresponding to Arabidopsis on the reference genome of Arabidopsis Araport 11 [15], while the reads from Alternaria were mapped on the Alternaria Abra43 reference genome [16]. Transcript abundance was also measured as transcripts per million (TPM) using Salmon [17].

DESeq2 algorithms (package version 1.36.0) [18] were used to obtain genes expression changes for both the plants and fungi. Pairwise comparisons for gene expression values were made between healthy and infected mutant seed (Table S1), as well as between the mutants and the WT infected by Alternaria (Tables S2 and S3). Genes with log2FC > 1 or <−1 and Benjamani–Hochberg adjusted p-values < 0.05 were considered as differentially expressed. Functional enrichment analysis was conducted using the ClusterProfiler 4.0 package version 4.12.6 [19]. To identify significantly enriched GO terms associated with biological processes, an adjusted p-value threshold of <0.05 was set following the Benjamani–Hochberg multiple testing correction [19].

To illustrate gene expression changes and DEGs, Venn diagrams [21] (Figure 1), stacked bar charts (Figure 3), and volcano plots (Figure 4) were employed. Some of the plots were generated using the ggplot2 (version 3.4.0) [24] package in R studio version 4.2.0.

4. User Notes

Previously, we provided an RNA dataset for the WT genotype, which exhibited specific expression patterns in response to Alternaria [9]. The transcriptomic analysis revealed a significant upregulation of indole metabolism in seeds infected by Alternaria [8]. These findings were further supported by a phenotyping study using defense-deficient mutants. Surprisingly, seeds from glucosinolate mutants infected by Alternaria showed a lower infection rate compared to the WT strain.

In line with our previous advancements, this dataset represents a valuable addition, providing detailed genetic expression information from glucosinolate mutant seeds exposed to Alternaria. It establishes a solid foundation for future research aimed at unraveling the synergistic function of glucosinolate pathways in seed defense and exploring how these interactions can be manipulated to enhance pathogen resistance in crops. Furthermore, this study encourages the analysis of transcriptomic data from both the host and the seed-borne pathogen perspectives, highlighting the significance of model pathosystem. These data can not only be leveraged in future research but can also form the basis for developing crop protection strategies and innovative disease management in sustainable agriculture.