High-throughput time series expression profiling of Plasmopara halstedii infecting Helianthus annuus reveals conserved sequence motifs upstream of co-expressed genes
BMC Genomics volume 24, Article number: 140 (2023)
Downy mildew disease of sunflower, caused by the obligate biotrophic oomycete Plasmopara halstedii, can have significant economic impact on sunflower cultivation. Using high-throughput whole transcriptome sequencing, four developmental phases in 16 time-points of Pl. halstedii infecting Helianthus annuus were investigated. With the aim of identifying potential functional and regulatory motifs upstream of co-expressed genes, time-series derived gene expression profiles were clustered based on their time-course similarity, and their upstream regulatory gene sequences were analyzed here. Several conserved motifs were found upstream of co-expressed genes, which might be involved in binding specific transcription factors. Such motifs were also found associated with virulence related genes, and could be studied on a genetically tractable model to clarify, if these are involved in regulating different stages of pathogenesis.
Plant pathogens have a diverse range of intracellular and extracellular effectors for host colonization and nutrient acquisition . Pathogens secrete effector proteins to evade the host immune system and to obtain nutrients from host tissue. Biotrophic pathogens acquire nutrients from the living host, while necrotrophic pathogens kill host cells in order to access nutrients . Oomycetes are fungus-like, eukaryotic microbes closely related to heterokont algae, i.e. the Phaeophyta, Xanthophyta, Chrysophyta, and Bacillariophyta [2,3,4]. Third to viruses and Fungi, oomycetes are the most important group responsible for plant diseases . The majority of notorious plant pathogenic oomycetes belong to the genus Phytophthora (over 200 species) and to the downy mildews (over 700 species) [4, 6].
The highest species diversity of oomycetes is in the hemi-biotrophic and biotrophic parasites of flowering plants, including several important crops . Obligate biotrophs, such as Plasmopara halstedii maintain a long-lasting interaction with their hosts [1, 8]. The genome assemblies of Pl. halstedii and H. annuus available in public repositories are 75.3 Mb and 3.6 Gb in size, and consist of 15,707 and 83,177 gene transcripts (both coding and non-coding genes), respectively [9, 10]. The sunflower downy mildew pathogen, Pl. halstedii, proliferates in Helianthus annuus as vegetative intercellular hyphae, which colonize the host leading to symptoms, i.e. systemic infection leads to dwarfing and chlorosis of leaves, and infected flowers produce abnormal infertile seeds, leading to oil yield losses [8, 11].
Transcriptional regulation impacts cellular processes in regulating cellular mRNA levels during developmental stages [12, 13]. The most well-understood control of gene expression in eukaryotes is the regulation of mRNA transcription with RNA polymerase II by regulatory proteins called transcription factors . Sequence-specific transcription factor (TF) proteins bind with their transcription factor binding sites (TFBS), affecting the binding affinity of RNA polymerase upstream of the coding sequence .
In eukaryotes, each gene has several functional elements, e.g. core promoter, proximal and distal promoter sequences, as well as regulatory motifs in the 5’UTR close to the transcriptional start, each with specific sequence motifs recognized by sequence-specific TF protein families . However, this knowledge is mainly derived from animals, yeast, and flowering plants and little is known about the genetic basis of transcriptional regulation in oomycetes. Most of the work has been focused on the core promoter sequence sites at which the transcription initiation complex is assembled [17, 18]. The core promoter consists of two important elements, a TATA box or another motif of similar function located 25–30 nucleotide upstream of the TSS and the transcription start site (TSS) or INR or “the +1 position” that is often represented by YYANWYY in humans and TCAGTY in Drosophila melanogaster, even though it should be noted that in humans, only a minority of promoters contain a TATA-box or other simple motifs, necessitating detailed bioinformatics analyses for determining their structure and function . In oomycetes, T-rich conserved motifs with a probably TATA-box like function are located 50–100 nucleotides upstream of start codon . Very little is known about the genetic basis of transcriptional regulation in oomycetes. Whole-genome analysis revealed the presence of extended FPR (MWTTTNC) motif surrounding the INR (YCAYTYY) or TSS, AT-rich represent functional non-canonical TATA-like elements with weak positional preference and DPEP motif (SAASMMS) located 25 nucleotides downstream of the TSS [21,22,23]. In an earlier study the genome of the sunflower pathogen Pl. halstedii has been sequenced  and a positional bias of INR + FPR supra-motifs and CCAAT within oomycete promoters has been found [10, 24]. Another study identified the conserved core effectors in 17 Pl. halstedii pathotypes potentially targeting the host organelles and thereafter validated from inducing hypersensitive response in host sunflower broad spectrum resistant lines . Studies in Plasmodium falciparum have shown that strong motif groups can be identified in the upstream regions (< 2000 nt of translational start site) of 13 sets of genes with a similar expression pattern suggesting functional coordination .
Illumina-based RNA sequencing technologies provide quantitative gene expression data and, thus, a basis for studying transcriptional regulation by grouping genes with similar expression patterns [27, 28]. Chromatin immunoprecipitation sequencing (ChIP-seq) experiments are a common method for identifying binding sites and their specific transcription factors, but ChIP-seq has limited potential for the identification of novel transcription factors as they rely on TFs that are not sequence-specific [29, 30]. Thus, RNA-seq might be a solution to identify novel regulatory sequences by associating TFBS with their downstream genes. Clustering algorithms can be used to identify gene expression changes over time, as similar expression patterns are likely to be the result of regulation by the same transcription factor family . An example using this approach is a recent study on Listeria monocytogenes . The study based on RNA-seq data from L. monocytogenes found a previously unreported partially palindromic motif M58.7 in the promoter regions of ribosomal protein encoding genes that regulate the growth of bacteria . De novo DNA motif discovery analysis in the late blight pathogen Ph. infestans and two other Phytophthora species (Ph. ramorum and Ph. sojae) identified 19 conserved DNA motifs that correlated with gene expression levels during infection . In another study, gene-expression changes in both H. arabidopsidis and its host Arabidopsis thaliana during infection were investigated in compatible and incompatible interactions, and identified some promoter elements associated with H. arabidopsidis genes . However, the studies only involved a few time points and were not focused on a genome-wide analysis of transcriptional changes.
In this study, the transcriptional changes during the infection of sunflower with Pl. halstedii were investigated by analyzing shifts in mRNA expression profiles during developmental transitions (zoospore release, establishment of infection, colonization, and induction of sporulation). The aim of the study was to investigate the hypothesis that the upstream regulatory sequences of co-expressed genes might have discernable overrepresented motifs in core promoter regions close to the coding sequences. For this, we chose to analyse the motif-rich stretch 100 bp upstream of genes that had been identified as containing the core promoter in previous studies, even though we are aware that many promoters will not be fully covered by this approach and that longer upstream sequences would likely have revealed additional motifs.
Experimental design, materials and methods
The study complies with relevant institutional, national, and international guidelines and legislation for plant ethics.
RNA isolation, library preparation and transcriptome sequencing
Approximately 250 sunflower seeds (Helianthus annuus, cultivar Giganteus) were moistened for 36 h, peeled, and potted into gardening soil (Gärtnerpflanzerde, Hornbach, Germany). The plants were grown at 16 °C until the cotyledons were fully developed, carefully taken out of the soil maintaining the roots, and washed with distilled water twice to remove residual attached soil. The plants were kept overnight in a moist chamber. In parallel, sporulation of Plasmopara halstedii was induced in six systemically infected plants by placing them in a moist chamber into darkness overnight at 12 °C.
Infection of seedlings and zoospore release phase (time points 5 min, 15 min)
The sporulating plants were removed from the moist chamber, left to dry for about 45 min before the sporangium solution was prepared by cutting the leaves off and thoroughly rinsing them in approximately 2 L of autoclaved, distilled water. Triplicate samples were taken after 5 min and 15 min by transferring 25 ml of the inoculation solution in a 50 mL tube and immediately adding 25 mL of 2-propanol (100%). The tube was centrifuged for 1 min at 6000 g. The supernatant was removed until 5 mL remained, the residual sample was resuspended and transferred into a new 15 mL tube to which an equal amount of RNAlater (ThermoFisher Scientific, Waltham, USA) was added. The tube was centrifuged as described above and the sporangia that accumulated in the interphase between RNAlater and 2-propanol were carefully pipetted off and transferred to a new 1.8 mL cryotube, which was filled up with RNAlater. All tubes were placed into the − 80 °C freezer immediately after sample collection was completed. After 60 min the rinsed seedlings were added to the sporangia solution and incubated in darkness at 16 °C.
Establishment of infection (time points 4 h, 8 h, 12 h, 24 h, 48 h, 72 h)
The time points covering the establishment phase were sampled by scraping off the surface of the cotyledons to enrich material from the pathogen. To ensure RNA integrity, the cotyledons were dunked into RNAlater™ solution and the epidermal layer of the cotyledons were immediately scraped off from the upper- and lower-leaf surface with a scalpel blade. The material was then transferred to a 1.8 mL cryotube containing 1 mL of RNAlater. In addition, the scalpel blade moistened with RNAlater before, after, and in-between the scraping of every cotyledon. In total four plants (with two cotyledons each) were scraped per sample, and three independent samples were taken.
After 8 h, the remaining seedlings were taken out from the inoculation solution and placed in a closed moistened plastic box. After 48 h all seedlings, except for those to be sampled at time-point 72 h, were potted into fresh gardening soil in a climate chamber of the mesocosm hall (Senckenberg Biodiversity and Climate Research Centre, Frankfurt, Germany), which was set to a daily rhythm with 14 h light at 16 °C and 70% relative humidity and 10 h darkness at 12 °C and the same humidity. The day/night cycle was inverted to the natural day/cycle so the plants were always sampled at the end of their day.
Colonization phase (time points 120 h, 222 h, 288 h)
Soil attached to the seedlings was rinsed off using autoclaved, deionized water. For each sample three seedlings were sampled by cutting a square of approximately 10 mm from the central part of each cotyledon with a sterile scalpel blade and transferring this to a 1.8 mL cryotube filled with RNAlater. The sample was immediately disrupted using a blunt pistil to ensure RNA preservation of the whole sample.
Induction of sporulation (time points 290 h, 292 h, 294 h, 296 h)
All residual plants not sampled at time-point 288 h, were cut, rinsed with autoclaved, distilled water and placed into a moist chamber in darkness to induce sporulation. Time-points 290 h and 292 h (induction of sporulation) were collected as described for the colonization phase but without rinsing the cotyledons, and sampling four plants instead of three per replicate. Time-points 294 h and 296 h (sporulation phase) were sampled similarly to the establishment phase by scraping off the epidermis with a sterile scalpel blade but only in areas with visible sporulation. Since visible sporulation was not only present on the cotyledons but also on the primary leaves, an additional three samples per replicate were taken at time point 296 h by scraping off the lower leaf surface of primary leaves that showed sporulation as described above.
The RNA was extracted from the RNAlater-preserved material using a NucleoSpin RNA Plant Mini Kit (Macherey-Nagel, Düren, Germany) according to the manufacturer’s instructions. RNAlater solution was removed prior extraction by filtration (samples with low amounts of tissue) or centrifugation (samples with high amounts of tissue). For filtration, the samples were filtered through the sterile filter of a filter-tip (Biozym, Hessisch Oldendorf, Germany), the filter removed aseptically, and the upper half millimetre of the filter with the desired material cut and homogenized using metal beads together with lysis buffer as described in the manual to the kit. Library preparation and transcriptome sequencing (illumina 100 bp paired end) was conducted by Eurofins Genomics (Luxemburg).
A flow-chart showing the methodology followed to prepare RNA sequencing data and to conduct downstream data analysis is given in Fig. 1.
Read preprocessing, quantification, normalization
The raw paired-end RNA-seq reads were processed using Trimmomatic . Standard sequencing primer and adapter sequences (parameters ILLUMINACLIP: TruSeq3-PE. fa:2:30:10:8:true) were trimmed and reads with an average Phred score less than 33 (−phred 33) and reads shorter than 20 nt (SLIDINGWINDOW:5:20 LEADING:3 TRAILING:3 MINLEN:30)were removed. The STAR mapper  was used for mapping the processed reads to both host (HanXRQr1.0, Ensembl Plants release 49 ;) and pathogen (Plasmopara_halstedii_gca_900000015.1, Ensembl Protists release 49 ;) gene transcripts including coding and non-coding genes. Initially, the reference index for the genomes was generated using the STAR mapper and later sequenced reads were mapped using the reference index. To perform quantification analyses, filtered reads were analyzed by FeatureCounts  of the Subread package v2.0.1. FeatureCounts accumulates gene count information by mapping reads to the exonic regions of genes (−t exon) and the parameter for reads overlapping in multiple transcripts were selected. Reads that mapped uniquely to the pathogen genome were used for downstream analyses and only locations with 10-fold or higher coverage were retained. As part of expression normalization, within-sample and between-sample normalization was performed across all 16 time points and their three biological replicates, each.
Differential gene expression analysis
To account for over-dispersion or high variability of genes, negative binomial distribution for each gene and gene-wise dispersions were estimated. Expression levels and Z-scores were calculated across all 48 samples (16 time-points) by subtracting the observed value from the mean expression level of the time-point and dividing this by the standard deviation following pairwise comparisons between time-points . The DESeq2 likelihood ratio test was conducted using False Discovery Rate corrected p-values or adjusted p-values (p-adj) ≤ 0.01 to identify highly significant differentially expressed genes between the time-points .
Expression profile pattern-based gene clustering
For gene clustering and visualization, after transformation hierarchical clustering was used to classify sets of genes with similar expression pattern on the basis of pairwise Kendall rank correlations (at least 15 genes per cluster, correlation > 0.7, testing the hypothesis of correlation for a confidence interval of 95%), and outliers from the cluster distribution were removed. For every clustered gene set, expression values are visualized in terms of Z-scores using DEGreport across the 16 time-points .
5’UTR dataset and background model
With the aim to search for over-represented regulatory motifs in the upstream regions of each cluster of genes, upstream regions (100 nt) were extracted using python scripts from the reference genome of Pl. halstedii  and a core promoter sequence dataset was generated. All extracted sequences were oriented in 5-prime to 3-prime direction. Using the fasta-get-markov Perl script in the MEME (Multiple Expression Motifs [EM] for Motif Elicitation) suite 5.3.0, a 3rd order Markov background model file from all 5’UTR regions (core promotor regions, e.g. the stretch up to 100 bp from the start codon) FASTA sequences was generated .
With the purpose of motif identification in core promoter regions located within 100 bp upstream of genes, MEME was used for clusters with less than 50 genes and STREME for clusters with more than 50 genes . Motif lengths of 4 to 18 nt (e-value ≤0.5) were searched for, using the zero or one occurrence per sequence model in the sequences of each cluster. The binomial test based on differential enrichment was set to search for motifs with a p-value ≤0.05, with up to 20 iterations on the reference genome of Pl. halstedii. To each set of gene clusters and reference upstream regions, an empirical 3rd order background Markov model was applied, based on all upstream regions of the Pl. halstedii genome and accounting for single-, di-, tri- and tetra-nucleotide distributions for each set. The discovered motifs in the clustered genes were searched against the reference genome motifs database of Pl. halstedii generated by STREME. AME (Analysis of motif enrichment) of the MEME suite was used to detect whether the identified potential transcription binding site motifs were enriched (e-value ≤0.5) in set(s) of the upstream sequences, as defined by belonging to the same expression cluster, compared to the reference genome motif sequences . To scan the sequences for individual matches to each of the enriched motif(s), FIMO (Find Individual Motif Occurrences) was used .
Inference of biological roles associated with enriched motifs
From the published genome of Pl. halstedii, genes annotated as pathogenicity related and those involved in core biological pathways were obtained from the supplementary data of Sharma et al. . Furthermore, classifications based on the published functional annotations were done using shell scripts. Genes highly expressed, as inferred from the RNA-seq data, and annotated as encoding for transcription factor (TF)-proteins were compared to those in the list of known predicted transcription factors in oomycetes as represented in FungiDB .
RNA-seq data pre-processing, mapping and identification of Pl. halstedii encoding transcripts
To infer differential expression during different infection stages of the obligate biotroph Plasmopara halstedii interacting with Helianthus annuus, a time-series analysis was done. This included zoospore formation (harvested zoosporangia; 5 min and 15 min after incubation in water), establishment of the infection on the cotyledons surface (IF; 4 h, 8 h, 12 h, 24 h, 48 h, 72 h), colonization phase (CO; 120 h, 221 h, 288 h) and sporulation phase (SP; 290 h, 292 h, 294 h, 296 h, sporulating primary leaves at 296 h) (Fig. 1). Forty-eight cDNA samples (three biological replicates per time-point) were done from the different stages of infection. The sequencing of the resulting 48 Illumina paired-end libraries resulted in 28.3 million to 34.6 million reads for each time-point. After filtering, 94.3 to 96.7% reads were used for downstream analyses (Table 1). These were mapped to the 15,707 predicted genes of Pl. halstedii and the 61,131 predicted genes of H. annuus using the splice-aware mapper STAR, and reads mapped to the host were excluded from the further analyses of this study (Additional files 1 and 2). The reads uniquely mapping to Pl. halstedii ranged from 0.19 million to 30.3 million. Of the 15,707 unigenes of Pl. halstedii 12,801 showed expression during the stages sampled (Additional file 3).
Data variability, normalization and statistical analysis of differential gene expression
Some genes had a higher variance than others, consequently their expression values varied more from one time-point to other. Differently expressed genes have been identified with the assumption of equal dispersion across all the time-points . With mean and variance linked by local regression, the negative binomial model fits best to the pathogen count data as their mean is less than their variance. To account for variance bias between and within samples, scaling according to library size was used as a first normalization step (Fig. 2, Additional file 3).
After additional removal of gene data with zero or low expression change (log fold change < 0.58, or fold change 1.5) a likelihood ratio test with 15 pairwise comparisons of 16 time points 11,689 genes were found to be significantly differentially expressed (adjusted p-value < 0.01), which is about 90% of the overall genes expressed.
Gene expression clustering
For identifying patterns among the 16 time points from the four infection phases, hierarchical clustering of the genes based on their transformed expression fold-change was done. Clusters of genes were generated by hierarchical clustering of a gene to gene expression distance matrix and separated using divisive coefficients of the clustering. Based on pairwise correlations and scaled expression, a total of 97 significant clusters with a minimum number of 15 genes per cluster, containing 8444 differentially expressed genes (Figs. 2c and 3, Additional file 3) were obtained.
Core promoter motifs in the Pl. halstedii genome
For investigating putative core promoter reference motifs, the upstream 100 nt from the start codon ATG were used for all genes of Pl. halstedii. By this approach, 17 over-represented, statistically significant motifs (p < 0.05) were found. Except for streme-4 (TCATTYTYMAATTTK, 990 occurrences), streme-10 (GCCAAYCA; 209 occurrences), streme-13 (CAGCAACATC, 3114 occurrences), to the best of our knowledge, other motifs have not been reported previously from downy mildews. Among the novel motifs and streme-2 (CTTCCTCTTC, 1181 occurrences) had the highest occurrence as compared to the other 15 significant motifs (Table 2). Notably, there were no clusters identified representing the streme-2 (CTTCCTCTTC, 1181 occurrences), streme-14 (CACATCAA, 405 occurrences), and streme-16 (CGTCGACCCTGAYCC, 34 occurrences) motifs (Additional file 4), suggesting a more general function.
Core promoter region analysis of co-regulated and/or co-related genes within clusters and motif discovery
The 97 clusters of co-regulated genes contained 8432 genes and 16 to 1158 members. For an efficient motif discovery, the clusters were further divided into two categories based on their size. STREME, based on an advanced ab initio motif discovery algorithm, generates more accurate motifs for clusters with at least 50 genes . Of the conserved motifs found using STREME on clustered gene data some matched with 17 known regulatory motifs. Based on de novo motif discovery in smaller clusters using MEME, 61 statistically significant motif(s) were found for 53 out of 97 clusters varying from 4 nt to 16 nt (Table 2; Additional file 5).
Pathogenicity related genes
Some highly supported motifs were found for clusters which are highly enriched with genes related to pathogenicity, exocytosis and vesicle transport, ion channels and calcium binding proteins, plant cell wall degrading enzymes and transcription factors. These are shown in Fig. 4, along with their expression profiles.
In total, 381 putative secreted effector encoding genes were expressed. Out of these 175 were annotated as RxLR effector candidates. In addition to this, 27 CRNs (Crinkler gene family), 103 secreted proteases, 13 protease inhibitors, and two genes coding for ATP-binding cassette (ABC) transporters were found in clusters with a similar expression pattern. The majority of the secreted proteases identified were serine-, aspartyl-, cysteine-, ubiquitin- and metallo-proteases (Table 3; Additional file 6).
Exocytosis and vesicle transport
Vesicle exocytosis, the fusion of a pathogen derived vesicle with the cell plasma membrane, releases vesicular contents to alter host immune responses . In this study 37 genes involved in exocytosis and protein export from the endoplasmic reticulum to the Golgi apparatus were found to be differentially expressed during different stages of infection, including SEC1, SEC3, SEC4, SEC13, SEC14, SEC16, SEC23, SEC24, SEC61, some SNARE proteins, and Rab GTPases (Table 4; Additional file 7).
Ion channels and calcium-binding proteins
Ion channels are specialized proteins, through which Na+, K+, Ca2+ and Cl− ions are transported through membranes. They are possibly important countering early host signaling to induce immune responses, including apoptosis or programmed cell death . There are in total 53 genes encoding for calcium-activated potassium channels, chloride channels, and voltage-gated ion channels differentially expressed throughout different lifecycle stages of the pathogen (Table 4; Additional file 8). Others differentially expressed ion channels include inward rectifier potassium channels (IRK), polycystin, and a sodium potassium-transporting ATPase subunit.
Calcium binding proteins (CBP) also act as Ca2+ transporters across cell membranes, thereby controlling Ca2+ concentration in the cytosol and regulating cellular functions. In response to a Ca2+ increase in the cell, CBP such as calmodulin and calcineurin indirectly regulate phosphorylation/dephosphorylation of transcription factors . Most of the differentially expressed calcium binding genes (33 genes) transcribed were classified as calcineurin, calmodulin, EF-H superfamily, CDKs, dynamin and protease kinase inhibitors. (Table 4; Additional file 8). If the differentiation of the mostly endogenously-acting proteins mentioned above are needed for pathogenicity and serve homoeostasis throughout different lifecycle stages needs to be clarified in future studies.
Plant cell wall degrading enzymes (PCWDEs)
Pathogens invade the host plant through the degradation of plant cell walls composed of polysaccharide components such as pectin, cellulose and hemicellulose . In the current study, 27 potential genes for secreted carbohydrate-active enzymes (CAZyme) were found differentially expressed, including members of carbohydrate esterase (CE), glycoside hydrolase (GH), glycosyl transferase (GT) and polysaccharide lyase (PL) families (Table 4; Additional file 9).
Transcription factors (TFs) or sequence-specific DNA binding factors are the proteins that bind to the upstream regulatory elements of genes and stimulate or inhibit transcription. In the present study, 135 transcription factors were found to be differentially expressed during different life-cycle stages, primarily from the myb factor and zinc finger families. In oomycetes, the myb DNA-binding domain is involved in growth and developmental processes , and 23 myb TFs were found to be differentially expressed in this study. In oomycetes, the zinc finger proteins form one of the largest families of transcriptional regulators of growth, development and pathogenesis [51, 52]. In this study, 74 members of zinc finger and 18 members of the bZIP families were found to be differentially expressed. Other families found were helix-turn-helix (HTH; 5), general transcription factors (10), homeobox (3), and heat shock related (2) TFs (Table 4; Additional file 10).
Despite of recent insights into core promotors of oomycetes [10, 22], clues for motifs involved in regulation are largely lacking, with a few exceptions [33, 53]. In this study a systems-level exploration of the of the differential transcription of genes expressed throughout the lifecycle of Plasmopara halstedii was done. Overall, 97 clusters of similarly regulated genes were observed, which might be useful to understand the functionality of the un-annotated genes from other annotated genes in the same expression cluster .
There is evidence for differential regulation of long non-coding RNAs (lncRNAs) in Ph. sojae . Relating to Jashni et al. , the current Pl. halstedii transcriptome study provides additional support to the presence of putative ncRNAs expressed at lower levels (28.5%) than coding RNAs. For the non-coding gene ENSRNAG00050153322 (U6atac minor spliceosomal RNA), the motif streme-12 (AACTTCAAC; p-value 6.37E-05; cluster 34) in the upstream DNA sequence has been found. This ncRNA was co-expressed with a gene cluster enriched with secreted crinklers, e.g. gene CEG41152 (CRN-like protein; LFLAK- Position: LYLAK-40), suggesting a potential function associated with effector regulation .
Prominent motifs and in oomycete core promotors
The candidate regulatory motifs streme-4 (TCATTYTYMAATTTK; 15 nt) and streme-10 (GCCAAYCA; 8 nt) bear similarities to a combined INR (YCAYTYY) + FPR (MWTTTNC) and the CCAAT box, respectively, which were detected in the oomycetes Ph. infestans, H. arabidopsidis, Ph. sojae, Globisporangium ultimum, and Saprolegnia parasitica [21, 57] and, thus, seem to be general motifs involved in gene expression and transcriptional start in oomycetes. In the current study, the combined INR-FPR motif was detected 990 times (enrichment p-value 0.000037) in 33 clusters of the whole set (Additional file 4), while streme-10 (putative CCAAT box) was detected 209 times (enrichment p-value 0.028) in 9 clusters, suggesting a low conservation of the motif. The putative regulatory motif streme-4 (TCATTYTYMAATTTK; 15 nt) and streme-13 (CAGCAACATC; 10 nt), to some extent, shows similarity with predicted transcription start site (TSP1, flanking upstream of TSP2 respectively) in core promoters of PinifC1, PinifC2, and PinifC3 genes of Ph. infestans, as well as putative homologues from Ph. sojae and Ph. ramorum . The motif streme-5 (ATACCGGTAT; 10 nt) and motif-45 (TAATATWAC; 9 nt; TATA-like) were identified upstream of a set of 206 and 75 genes, respectively, which are enriched with RxLRs, proteases and transcription factors. These motifs correspond to motif-9 (GTACCGGTA; 9 nt) and motif-7 (TACTATTAGTA; 11 nt) respectively, found in the study of Seidl et al. .
Comparison with other eukaryotes
The 4–16 nt long motifs identified in this study for 53 clusters occur in 2.15 to 100% of the genes of the corresponding clusters. In total, 61 cluster-unique motifs were identified using MEME and STREME (Additional files 4 and 5). In relation to expression-based cluster motif analysis performed in Plasmodium falciparum , the motif streme-13 (CAGCAACATC; 10 nt) found in this study was observed to be similar with CAAC-motif sets identified in the upstream regions of the DNA replication machinery set in Pd. falciparum. Additionally, the C-rich motif streme-11 (ARCCCCGYTRCA; 12 nt) identified in 60 sequences is similar to motifs (CCCCAT, TTCCC) upstream region of set of 15 mitochondrial genes in that species. The streme-1 motif (DRTGTGGWCCACAYH; 15 nt), which occurred in upstream region of 379 genes, is similar to the TGTG and CACA motif-sets, identified in the Pd. falciparum upstream regions of genes coding for genes involved in the organelle translation machinery and the proteasome. In this study, streme-6 (ATGGCTACA; 9 nt) and streme-12 (AACTTCAAC; 9 nt), were found to bear similarities with the most overrepresented motif E2F (TGGCGCCA; 8 nt) associated with DNA replication and glycolysis and a motif (KAACTA; 6 nt) of unknown function in the protist parasite Cryptosporidium parvum . In addition, the second-most overrepresented motif-13 (AAACGAGAAT; 10 nt) of cluster 14, located at 27 out of 218 sequences, had previously been reported as GAGA-motif. As reported by Vela-Corcía et al. , the motifs motif-8 (AAATCAAAAC; 10 nt), motif-10 (ATTGGCTRAAA; 11 nt), motif-26 (CTTTTGAC; 8 nt), motif-31 (ATGATTGGCT; 10 nt), as well as motif-32 (ATGATTGACGA; 11 nt) bear similarities with the TTGNC consensus sequence for σ70-family factors of validated promoters of the Pseudomonas aeruginosa PA14 genome.
A total of 381 putative pathogenicity-related genes, including 175 coding for putative RXLR effectors and 26 for putative Crinklers (CRNs) containing a secretion signal were identified in the RNA-seq gene clustering analysis. In a study on Plasmopara viticola, PvRxLR67 and some other effectors were upregulated 1 dpi, but were also active in later infection stages in a high virulence strain . In the current study, CEG37612 (cl 14; IF phase; log2FC 12.0; dEER-like motif position (DEI): 30), CEG45621 (cl 8; IF phase; log2FC 11.7; RxLR-like motif position (RNLV): 27; dEER like motif position (EER): 36), and CEG49100 (cl 14; IF phase; log2FC 9.3; RxLR-like motif position (KLLR): 40; dEER like motif position (EER): 52) were significantly higher expressed as compared comparison to other putative RxLR effector encoding genes, suggesting an essential role in pathogenicity that could be explored in future studies.
Among 27 secreted putative CRN effectors, two highly expressed Crinkler genes were CEG37824 (cl 24; SP phase; log2FC 3.5; HVLVVVP-position: HVLVVVP-514) and CEG47474 (cl 7; IF phase; log2FC 4.2; HVLVVVP-position: HVLVELP-189). During the infection establishment, gene CEG37265 (cl 60; IF phase; log2FC -5.3; LFLAK-Position: LYLAK-47) has been down regulated to the largest extent. Stam et al.  classified CRNs into two classes on the basis of expression patterns. Genes with no detectable expression in the early stages and increase after the infection establishment (class 2 according to Stam et al. ) are CEG40812 (cl 11; LFLAK Positions: LFLAK-14), CEG42045 (cl 2; LFLAK-Position: EFLAN-92; HVLVVVP-position: HVLVVVP-302), and CEG49843 (cl 25; HVLVVVP-position: HVLVVIP-104). This suggests, that these genes are probably important in maintaining the infection process after establishment.
Pathogens secrete proteases and protease inhibitors to manipulate host defense for a successful colonization . In this study 27% of the proteases were predicted to be secreted and, thus, have a potential role in pathogenicity. The genes found to be most up-regulated are CEG48424 (metalloprotease family) in cl 8, IF phase, log2FC 9.4, while CEG35751 (Cytosolic Ca2-dependent cysteine protease (calpain), large subunit (EF-Hand protein superfamily) in cl 11, IF phase, log2FC − 6.1 was most down regulated. The motifs streme-4 (TCATTYTYMAATTTK; 15 nt; putative INR + FPR) and streme-10 (GCCAAYCA; 8 nt; putative CCAAT box) regulating the genes transcription in cl 8 and cl 11 were found. This situation is similar to findings on proteolytic enzymes in the oomycete fish pathogen Sa. parasitica where seven and predicted secreted proteases were found to be expressed during infection [63, 64].
Furthermore, the secreted protease inhibitor coding genes highly up-regulated are CEG40179 (cl 29, SP phase, log2FC 7.87) and CEG36013 (cl 14, IF phase, log2FC 11.6). The expression patterns of the genes are in opposite direction to each other, even though they share a common motif streme-4 (TCATTYTYMAATTTK; 15 nt; putative INR + FPR) suggesting that other TF binding sites are more important for the regulation of the genes.
Only two genes encoding an ATP-binding cassette (ABC) transporter were identified in the clusters, CEG48435 (ABC transporter ATP-binding protein) and CEG37981 (ABC transporter e family member 2), in cl 12 and cl 59, respectively. CEG48435 and CEG37981 genes were increasingly up-regulated during infection, to the maximum extent in the late infection stages and expression down regulated on the induction of sporulation. The upstream motifs streme-1 (DRTGTGGWCCACAYH; 15 nt; TGTG+CACA like), streme-3 (GATTGGHTRAAAWDW; 15 nt), streme-4 (TCATTYTYMAATTTK; 15 nt; putative INR + FPR), streme-10 GCCAAYCA; 8 nt; putative CCAAT box) have been found common in cluster 12 and cl 59. There are recent findings that suggest these transporters might be involved in the detoxification of phytoalexins [65, 66].
In the beginning of the infection (IF) phase the voltage-gated ion channels identified are generally downregulated but upregulated during CO and SP phase. An example of this is CEG38124 (Chloride channel family) in cl 11; CO phase; log2FC 5.94, which has been suggested to have importance in cell volume regulation  and, thus, might be an important component of plasma extension leading to sporulation.
Plant cell wall degrading enzymes (PCWDEs)
Oomycetes express enzymes capable to dissolve or soften plant primary cell walls for initial penetration and the formation of haustoria. CAZyme coding genes were reported in, e.g. Phytophthora and Plasmopara viticola [68, 69]. A total of 27 genes coding for secreted carbohydrate-active enzymes (CAZyme) were found to be significantly upregulated at some stages of infection. These included carbohydrate esterases, glycoside hydrolases, glycosyl transferases and polysaccharide lyases (Additional file 9). In this study, the highly up-regulated and secreted-protein coding gene CEG46932 (a glycoside hydrolase family 6 protein) in cl 6 (IF phase; log2FC 9.5) had its maximum expression in the beginning of the infection phase (Additional files 5 and 9). This expression profile indicated a potential role in host penetration by degradation or modification of polymerous carbohydrates within the host cell wall.
The strong upregulation of some genes coding for transcription factors, e.g. CEG35443 (myb-like DNA-binding) in cl 1, IF phase, log2FC7.8 and CEG42655 (Transcription factor IIB) in cl 2, CO phase, log2FC 6.7 suggests their possible involvement as regulators of the infection and colonization processes, respectively. Further functional studies and validations are needed to elucidate the significance of these transcription factors in host-pathogen interaction.
Conclusions and outlook
High-throughput RNA sequencing is an important tool for linking gene expression and function with potential regulatory motifs. In this study 41 novel and 169 known motifs were found in clustered gene groups with 16 to 1158 members. Some motifs were observed from multiple clusters suggesting a more general role in the cellular processes of Pl.halstedii. To the best of our knowledge, the identified putative novel motifs and known motifs except for streme-4 (TCATTYTYMAATTTK; 15 nt; putative INR + FPR), streme-10 (GCCAAYCA; 8 nt; putative CCAAT box) and streme-12 (AACTTCAAC; 9 nt; CAAC-like), have not been reported previously from oomycetes. Interestingly, some motifs such as streme-1 (DRTGTGGWCCACAYH; 15 nt; TGTG+CACA like) represented condensations of previously identified motifs. This might imply targeting by two independent transcription factors, probably as a possibility for regulation by competitive binding. However, it is clear that the current study only provided a first glimpse into genes with putative virulence functions and the diversity of potential transcription factor binding sites identified by motif searchers. This will need to be followed up by laboratory experiments to verify the importance of the highly differentially expressed genes in pathogenicity and to identify the potential functional redundancy of highly similar motifs and the potential transcription factors binding to the various motifs found. In the context of host-pathogen interactions, a systems biology view for host pathogen interaction could be a powerful means for achieving a system-wide understanding of cellular processes involved in pathogenicity. However, it should be noted that in the current study, only the 100 nt upstream of genes were investigated for core promoter motifs to enable a broad screening for potential motifs. Many core promoters will not have been fully covered by this approach, and an extension to 500 bp or more upstream of genes would likely lead to the discovery of additional regulatory motifs.
Availability of data and materials
All data used in this study are available from GenBank under the accession number PRJEB49134 (https://www.ncbi.nlm.nih.gov/bioproject/49134).
differentially expressed genes
transcription factor binding site
Fawke S, Doumane M, Schornack S. Oomycete interactions with plants: infection strategies and resistance principles. Microbiol Mol Biol Rev. 2015;79:263–80. https://doi.org/10.1128/MMBR.00010-15.
Beakes GW, Sekimoto S. The evolutionary phylogeny of oomycetes-insights gained from studies of Holocarpic parasites of algae and invertebrates. In: Oomycete Genetics and Genomics: Diversity, Interactions, and Research Tools 2008.
Burki F, Shalchian-Tabrizi K, Minge M, Skjæveland Å, Nikolaev SI, Jakobsen KS, et al. Phylogenomics reshuffles the eukaryotic supergroups. PLoS One. 2007;2(8):e790.
Thines M. Phylogeny and evolution of plant pathogenic oomycetes-a global overview. Eur J Plant Pathol. 2014;138(3):431–47.
Anderson PK, Cunningham AA, Patel NG, Morales FJ, Epstein PR, Daszak P. Emerging infectious diseases of plants: pathogen pollution, climate change and agrotechnology drivers. Trends Ecol Evol. 2004;19(10):535–44.
Beakes GW, Honda D, Thines M. Systematics of the Strami- nipila: Labyrinthulomycota, Hyphochytriomycota, and Oomycota. New York: Syst Evol D J McLaughlin J Spat Springer; 2014. p. 39–97.
Thines M, Choi YJ. Evolution, diversity, and taxonomy of the peronosporaceae, with focus on the genus peronospora. Phytopathology. 2016;106(1):6–18.
Gascuel Q, Martinez Y, Boniface MC, Vear F, Pichon M, Godiard L. The sunflower downy mildew pathogen Plasmopara halstedii. Mol Plant Pathol. 2015;16(2):109–22.
Kane NC, Gill N, King MG, Bowers JE, Berges H, Gouzy J, et al. Progress towards a reference genome for sunflower. Botany. 2011;89:429–37.
Sharma R, Xia X, Cano LM, Evangelisti E, Kemen E, Judelson H, et al. Genome analyses of the sunflower pathogen Plasmopara halstedii provide insights into effector evolution in downy mildews and phytophthora. BMC Genomics. 2015;16:741.
Döken MT. Plasmopara halstedii (Farl.) Berl. et de Toni in sunflower seeds and the role of infected seeds in producing plants with systemic symptoms. J Phytopathol. 1989;124:23–6. https://doi.org/10.1111/j.1439-0434.1989.tb04890.x.
Orphanides G, Reinberg D. A unified theory of gene expression. Cell. 2002;108(4):439–51.
Maniatis T, Reed R. An extensive network of coupling among gene expression machines. Nature. 2002;416:499–506.
Reményi A, Schöler HR, Wilmanns M. Combinatorial control of gene expression. Nat Struct Mol Biol. 2004;11:812–5. https://doi.org/10.1038/nsmb820.
Todeschini AL, Georges A, Veitia RA. Transcription factors: specific DNA binding and specific gene regulation. Trends Genet. 2014;30(6):211-9.
Tsonis PA. Transcription in eukaryotes. In: Anatomy of Gene Regulation. 2010.
Nakatani Y, Brenner M, Freese E. An RNA polymerase II promoter containing sequences upstream and downstream from the RNA startpoint that direct initiation of transcription from the same site. Proc Natl Acad Sci. 1990;87:4289–93. https://doi.org/10.1073/pnas.87.11.4289.
Kadonaga JT. Perspectives on the RNA polymerase II core promoter. Wiley Interdiscip Rev Dev Biol. 2012;1:40–51. https://doi.org/10.1002/wdev.21.
Umarov R, Kuwahara H, Li Y, Gao X, Solovyev V. Promoter analysis and prediction in the human genome using sequence-based deep learning models. Bioinformatics. 2019;35:2730–7.
Ghimire B, Saraiva M, Andersen CB, Gogoi A, Saleh M, Zic N, et al. Transformation systems, gene silencing and gene editing technologies in oomycetes. Fungal Biology Reviews. 2022;40:37–52.
McLeod A, Smart CD, Fry WE. Core promoter structure in the oomycete phytophthora infestans. Eukaryot Cell. 2004;3(1):91–9.
Roy S, Poidevin L, Jiang T, Judelson HS. Novel core promoter elements in the oomycete pathogen Phytophthora infestans and their influence on expression detected by genome-wide analysis. BMC Genomics. 2013;14:106.
Roy S, Kagda M, Judelson HS. Genome-wide prediction and functional validation of promoter motifs regulating gene expression in spore and infection stages of Phytophthora infestans. PLoS Pathog. 2013;9:e1003182. https://doi.org/10.1371/journal.ppat.1003182.
Janky R, van Helden J. Evaluation of phylogenetic footprint discovery for predicting bacterial cis-regulatory elements and revealing their evolution. BMC Bioinformatics. 2008;9:37.
Pecrix Y, Buendia L, Penouilh-Suzette C, Maréchaux M, Legrand L, Bouchez O, et al. Sunflower resistance to multiple downy mildew pathotypes revealed by recognition of conserved effectors of the oomycete Plasmopara halstedii. Plant J. 2019;97:730–48. https://doi.org/10.1111/tpj.14157.
Iengar P, Joshi NV. Identification of putative regulatory motifs in the upstream regions of co-expressed functional groups of genes in plasmodium falciparum. BMC Genomics. 2009;10:18.
Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 2009;10(1):57–63.
Jakt LM, Cao L, Cheah KSE, Smith DK. Assessing clusters and motifs from gene expression data. Genome Res. 2001;11(1):112–23.
Karimzadeh M, Hoffman MM. Virtual ChIP-seq: Predicting transcription factor binding by learning from the transcriptome. bioRxiv. 2018;23(1):126.
Park PJ. ChIP-seq: advantages and challenges of a maturing technology. Nat Rev Genet. 2009;10(10):669–80.
Song J, Bjarnason J, Surette MG. The identification of functional motifs in temporal gene expression analysis. Evol Bioinforma. 2005;1:84–96.
Sultan I, Fromion V, Schbath S, Nicolas P. Statistical modelling of bacterial promoter sequences for regulatory motif discovery with the help of transcriptome data: Application to Listeria monocytogenes. J R Soc Interface. 2020;7(171):20200600.
Seidl MF, Wang RP, van den Ackerveken G, Govers F, Snel B. Bioinformatic inference of specific and general transcription factor binding sites in the plant pathogen Phytophthora infestans. PLoS One. 2012;7(12):e51295.
Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20.
Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21.
Yang L, Smyth Gordon K, Wei S. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–30.
Tavazoie S, Hughes JD, Campbell MJ, Cho RJ, Church GM. Systematic determination of genetic network architecture. Nat Genet. 1999;22(3):281–5.
Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:550.
Pantano L. DEGreport: Report of DEG analysis. 2020. https://doi.org/10.18129/B9.bioc.DEGreport.
Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, et al. MEME suite: tools for motif discovery and searching. Nucleic Acids Res. 2009;37(Web Server issue):W202–8.
Bailey TL. STREME: Accurate and versatile sequence motif discovery. bioRxiv. 2020;11(23):394619.
McLeay RC, Bailey TL. Motif enrichment analysis: a unified framework and an evaluation on ChIP data. BMC Bioinformatics. 2010;11:165.
Grant CE, Bailey TL, Noble WS. FIMO: scanning for occurrences of a given motif. Bioinformatics. 2011;27(7):1017–8.
Basenko EY, Pulman JA, Shanmugasundram A, Harb OS, Crouch K, Starns D, et al. FungiDB: An integrated bioinformatic resource for fungi and oomycetes. J Fungi. 2018;4(1):39.
Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11:R106.
Leborgne-Castel N, Bouhidel K. Plasma membrane protein trafficking in plant-microbe interactions: a plant cell point of view. Front. Plant Sci. 2014;5:735.
Kurusu T, Mitsuka D, Yagi C, Kitahata N, Tsutsui T, Ueda T, et al. Involvement of S-type anion channels in disease resistance against an oomycete pathogen in Arabidopsis seedlings. Commun Integr Biol. 2018;11(3):1–6.
Ikura M, Osawa M, Ames JB. The role of calcium-binding proteins in the control of transcription: structure to function. BioEssays. 2002;24(7):625–36.
Kubicek CP, Starr TL, Glass NL. Plant cell wall-degrading enzymes and their secretion in plant-pathogenic fungi. Annu Rev Phytopathol. 2014;52:427–51.
Xiang Q, Judelson HS. Myb transcription factors in the oomycete phytophthora with novel diversified DNA-binding domains and developmental stage-specific expression. Gene. 2010;453(1-2):1–8.
Wang Y, Dou D, Wang X, Li A, Sheng Y, Hua C, et al. The PsCZF1 gene encoding a C2H2 zinc finger protein is required for growth, development and pathogenesis in Phytophthora sojae. Microb Pathog. 2009;47:78–86. https://doi.org/10.1016/j.micpath.2009.04.013.
Gamboa-Meléndez H, Huerta AI, Judelson HS. bZIP transcription factors in the oomycete Phytophthora infestans with novel DNA-binding domains are involved in defense against oxidative stress. Eukaryot Cell. 2013;12(10):1403–12.
Tani S, Judelson H. Activation of zoosporogenesis-specific genes in Phytophthora infestans involves a 7-nucleotide promoter motif and cold-induced membrane rigidity. Eukaryot Cell. 2006;5:745–52. https://doi.org/10.1128/EC.5.4.745-752.2006.
Castro NC, Azevedo PJ. Significant motifs in time series. Stat Anal Data Min. 2012;5(1):35–53.
Wang Y, Ye W, Wang Y. Genome-wide identification of long non-coding RNAs suggests a potential association with effector gene transcription in Phytophthora sojae. Mol. Plant Pathol. 2018;19(9):2177–86.
Rujirawat T, Patumcharoenpol P, Kittichotirat W, Krajaejun T. Oomycete gene table: An online database for comparative genomic analyses of the oomycete microorganisms. Database. 2019;2019:baz082.
Asai S, Rallapalli G, Piquerez SJM, Caillaud M-C, Furzer OJ, Ishaque N, et al. Expression profiling during Arabidopsis/downy mildew interaction reveals a highly-expressed effector that attenuates responses to salicylic acid. PLoS Pathog. 2014;10:e1004443. https://doi.org/10.1371/journal.ppat.1004443.
Oberstaller J, Joseph SJ, Kissinger JC. Genome-wide upstream motif analysis of Cryptosporidium parvum genes clustered by expression profile. BMC Genomics. 2013;14:516. https://doi.org/10.1186/1471-2164-14-516.
Potvin E, Sanschagrin F, Levesque RC. Sigma factors in pseudomonas aeruginosa. FEMS Microbiol Rev. 2008;32:38–55. https://doi.org/10.1111/j.1574-6976.2007.00092.x.
Gómez-Zeledón J, Spring O. Up-regulated RxLR effector genes of plasmopara viticola in synchronized host-free stages and infected leaves of hosts with different susceptibility. Fungal Biol. 2018;122(12):1125–33.
Stam R, Jupe J, Howden AJM, Morris JA, Boevink PC, Hedley PE, et al. Correction: identification and characterisation CRN effectors in Phytophthora capsici shows modularity and functional diversity (PLoS ONE). PLoS One. 2013;8(3):e59517.
Jashni MK, Mehrabi R, Collemare J, Mesarich CH, de Wit PJGM. The battle in the apoplast: further insights into the roles of proteases and their inhibitors in plant–pathogen interactions. Front Plant Sci. 2015;6:584.
Phillips AJ, Anderson VL, Robertson EJ, Secombes CJ, van West P. New insights into animal pathogenic oomycetes. Trends Microbiol. 2008;16(1):13–9.
Torto-Alalibo T, Tian M, Gajendran K, Waugh ME, Van West P, Kamoun S. Expressed sequence tags from the oomycete fish pathogen saprolegnia parasitica reveal putative virulence factors. BMC Microbiol. 2005;5:46.
Vela-Corcía D, Aditya Srivastava D, Dafa-Berger A, Rotem N, Barda O, Levy M. MFS transporter from Botrytis cinerea provides tolerance to glucosinolate-breakdown products and is required for pathogenicity. Nat Commun. 2019;10:2886. https://doi.org/10.1038/s41467-019-10860-3.
Stefanato FL, Abou-Mansour E, Buchala A, Kretschmer M, Mosbach A, Hahn M, et al. The ABC transporter BcatrB from Botrytis cinerea exports camalexin and is a virulence factor on Arabidopsis thaliana. Plant J. 2009;58:499–510. https://doi.org/10.1111/j.1365-313X.2009.03794.x.
Jentsch TJ, Friedrich T, Schriever A, Yamada H. The CLC chloride channel family. Pflugers Archiv European. J Physiol. 1999;437(6):783–95.
Ospina-Giraldo MD, Griffith JG, Laird EW, Mingora C. The CAZyome of phytophthora spp.: a comprehensive analysis of the gene complement coding for carbohydrate-active enzymes in species of the genus Phytophthora. BMC Genomics. 2010;11:525.
Yin L, An Y, Qu J, Li X, Zhang Y, Dry I, et al. Genome sequence of Plasmopara viticola and insight into the pathogenic mechanism. Sci Rep. 2017;7:46553.
We thank Tahir Ali, Ann-Katrin Buch and Bagdevi Mishra for valuable suggestions on the study and Claus Weiland for assisting in server access.
Open Access funding enabled and organized by Projekt DEAL. SB was supported for carrying out research under the DAAD doctoral program. MT is supported by LOEWE in the framework of The Centre for Translational Biodiversity Genomics (TBG), funded by the government of Hessen. The funders had no role in study design, data collection and analysis, decision to submit the work for publication, or preparation of the manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Bharti, S., Ploch, S. & Thines, M. High-throughput time series expression profiling of Plasmopara halstedii infecting Helianthus annuus reveals conserved sequence motifs upstream of co-expressed genes. BMC Genomics 24, 140 (2023). https://doi.org/10.1186/s12864-023-09214-7