Northern Wild Rice (Zizania palustris L.) (NWR) is a diploid (2n=2x=30) outcrossing, aquatic, C3 Gramineae species that grows in shallow waters of rivers and lakes in North America (Oelke, 2007). The species belongs to the subtribe Zizaniinae (Xu et al., 2010), within the tribe Oryzeae, distinct from the subtribe Oryzinae, containing white rice (Oryza sativa L.) (Guo and Ge, 2005). Despite the subtribes diverging 26–30 million years ago, genomic synteny and collinearity between species of Zizania and Oryza are substantial (Haas et al., 2021; Yan et al., 2022). For example, the genomes of NWR, Z. latifolia, and Oryza spp., including O. sativa, share 14,120 protein-coding orthologous groups (Haas et al., 2021). This is a relevant framework to understand the evolutionary history of the NWR genome and guide research for NWR trait development, such as reduced seed-shattering, since microsyntenic gene blocks for shattering between this species and O. sativa have been found (Haas et al., 2021). In addition, microsynteny of genomic areas for genes putatively involved in phytoalexin production is present among NWR, Z. latifolia, and O. sativa (Yan et al., 2022; Wu et al., 2011; this study), which could facilitate future functional gene studies. In addition, genomic comparisons can lead to a better understanding of vulnerabilities since NWR and O. sativa share susceptibility to some of the same pathogens (Lee, 1992; Samac and Castell-Miller, 2023).

Since the 1950s, NWR has been commercially cultivated in flooded paddies (Oelke, 2007), where fungal diseases often reduce grain production (Samac and Castell-Miller, 2023). Fungal brown spot (FBS) (Johnson and Percich, 1992), caused by Bipolaris oryzae (Breda de Haan) Shoemaker [Cochliobolus miyabeanus (S. Ito & Kurib.) Drechsler ex Dastur], is one of the most devastating diseases, causing a serious outbreak in 2015 (Castell-Miller et al., 2021) ( Figure 1A ). B. oryzae is mostly considered a necrotrophic fungus (Condon et al., 2013); thus, it presumably kills plant tissue during or soon after invasion (Shao et al., 2021). In artificial inoculations of susceptible NWR cultivars, K2 and Johnson, the fungus has typical bipolar germination by 8 h after inoculation with initial single germ tubes that branch and develop club-shaped appressoria by 24 h. Mostly direct plant cuticle penetration occurs, followed by intra- and intercellular fungal growth. Lesions form between 18 h and 48 h after inoculation (Mitchell Schilkli, 1984). Typical FBS symptoms are purple to brownish spots that enlarge into dark oval or diamond-shaped lesions with tan necrotic centers and chlorotic halos (Johnson and Percich, 1992). FBS can develop on leaves, sheaths, stems, and panicles ( Figures 1B, C ) (Samac and Castell-Miller, 2023). To mitigate disease damage in commercial NWR paddies, an integrated disease management system is followed, based on the management of infected residue, plant nutrition, use of fungicides, and planting of improved disease resistant cultivars (Johnson and Percich, 1992; Porter, 2008; Castell-Miller et al., 2021).

B. oryzae also causes brown spot (BS) disease in white rice (Lee, 1992), where it can inflict significant grain losses (Sunder et al., 2014). While the genetic mechanisms for FBS resistance in NWR are currently unknown, several quantitative trait loci (QTLs) conferring white rice BS resistance have been identified. These QTLs explain roughly 9% to 23% (Mizobuchi et al., 2016), 23% to 26% (Matsumoto et al., 2017), and 15% to 20% (Ota et al., 2021) of the overall phenotypic variation.

Detection of pathogen-/microbe-/damage-associated molecular patterns (PAMP, MAMP, and DAMP, respectively) by pattern recognition receptors (PRRs) such as receptor-like kinases (RLKs) and receptor-like proteins (RLPs) (Zipfel, 2014; Buendia et al., 2018) results in PAMP-triggered immunity (PTI) (Jones and Dangl, 2006; Zipfel, 2009; Boutrot and Zipfel, 2017). PTI triggers production of reactive oxygen species (ROS), activation of ion channels and mitogen-activated protein kinase (MAPK) cascades, leading to transcriptional reprogramming that produces antimicrobial proteinaceous and non-protein-based molecules (Lo Presti et al., 2015; Saijo et al., 2018). In general, PTI confers quantitative resistance to necrotrophic pathogens lacking host specificity (Liao et al., 2022). PAMPs from necrotrophic fungi include chitin and derivatives, and endo-polygalacturonases (PGs), among other molecules (Wang et al., 2014). Fungi often avoid or overcome PTI by secreting apoplastic and symplastic effectors in a spatial and temporal fashion that can manipulate plant signaling and defense responses (Toruño et al., 2016; Buendia et al., 2018). Effectors are usually perceived through direct or indirect contact (Dodds and Rathjen, 2010) by resistance (R) proteins with conserved intracellular domains usually of the nucleotide-binding leucine-rich receptor (NB-LRR) class (Jones and Dangl, 2006; Lo Presti et al., 2015) activating a highly specific, rapid, and robust response called effector-triggered immunity (ETI) (Jones and Dangl, 2006; Dodds and Rathjen, 2010), often referred to as a hypersensitive reaction (HR). ETI conferring gene-to-gene resistance is an uncommon response to necrotrophic pathogens (Staal et al., 2008), and in some cases, effector perception by R proteins leads to susceptibility to host-specific necrotrophy that exploits plant defenses in an inverse gene-for-gene manner (Faris et al., 2010; Lorang, 2019). Plant defenses against necrotrophic pathogen threats are usually built independently of ETI (Liao et al., 2022).

Overall, effectors of necrotrophic pathogens are phytotoxins, either host selective (HS) or non-HS, including polyketides, non-ribosomal peptides, alkaloids, terpenes, and proteinaceous molecules that can cause cell death, and even induce cell survival during initial plant colonization and manipulate plant defenses (Toruño et al., 2016; Shao et al., 2021). Some HS toxins (HST) from necrotrophic fungi, including Cochliobolus spp., can elicit an HR to obtain nutrients from dead tissues (Wang et al., 2014; Liu et al., 2015; Lorang, 2019). Although B. oryzae isolates have genomes well-equipped with pathogenicity/virulence genes (Oide et al., 2006; Moriwaki et al., 2007; Condon et al., 2013; Castell-Miller et al., 2016), no HSTs have been found as in other Cochiobolus spp (Condon et al., 2013). Non-HST variants of sesquiterpene ophiobolins are produced by Bipolaris spp (Xiao et al., 1991) that mainly affect roots and induce stomatal opening (Au et al., 2000). Type A appears to be associated with the virulence of B. oryzae on rice (De Bruyne et al., 2016).

Little is known about the molecular basis of innate defenses employed by NWR against plant pathogens. Recently, it was reported that the Z. latifolia genome has clusters of genes putatively involved in phytoalexin biosynthesis (Yan et al., 2022).

In this study, we investigated the dual transcriptomes of the NWR cultivar Itasca–C12 (Porter, 2008), and the B. oryzae isolate TG12Lb2 (Castell-Miller et al., 2016) in the early stage of disease development since expression profiling at the onset of plant–pathogen interactions will likely identify plant responses associated with perception, defense activation, and fungal pathogenesis pathways (Shen et al., 2017). This decision was also based on observations that in NWR, FBS leaf symptoms develop between 18 h and 30 h in greenhouse conditions, depending on the plant and fungus genotypes used, with clearly visible lesions at 48 h along all the leaf areas inoculated.

This research will increase our understanding of this pathosystem, with the goal of identifying relevant candidate fungal virulence/effector genes for population functional studies, and importantly, it constitutes the baseline to uncover NWR defense responses as a primary step toward functional genomic association studies, development of molecular marker-assisted selection, and potentially pyramiding FBS-resistant alleles.

The specific objectives were to (1) assemble and annotate the non-redundant reference transcriptomes of NWR and B. oryzae and (2) identify NWR and B. oryzae genes putatively associated with defense and pathogenicity, respectively.

Bipolaris oryzae is one of the most consequential threats to the NWR grain production and germplasm continuity in natural and cultivated environments (Johnson and Percich, 1992; Samac and Castell-Miller, 2023). To expand our understanding of plant defense and fungal virulence mechanisms, we designed an RNA sequencing study using an improved disease-resistant cultivar, Itasca-C12 (Porter, 2008), and a virulent B. oryzae isolate TG12Lb2 (Castell-Miller et al., 2016), during the first 48 h of their interaction.

We assembled and annotated the non-redundant reference transcriptome of NWR in the absence and in the presence of the fungus at 24 h and 48 h after inoculation (t_WRim) and that of B. oryzae with genes expressed in vitro and in planta at the same time points (t_Boiv_ip). The t_WRim resulted in a total of 93,610 transcripts (N50 = 1,656 bp), which roughly represents 13.6% of those initially generated in the NWR whole genome transcriptome (WGT) (Haas et al., 2021). The larger length (over 172 bp) of N50 transcripts in this study could be attributed to a reduced subset of expressed genes from mock- and B. oryzae-infected leaves compared with those of the WGT that was based on mostly all plant organ tissues (Haas et al., 2021). The percentage of NWR transcript annotations in this study that matched reference NWR genes (~63%) was similar to that found during the genome annotation project (Haas et al., 2021). The current version of the draft NWR genome has some issues with resolution of a high percentage of repetitive sequences (76.4%). Additional challenges in our NWR transcript annotations could be attributed to the heterogeneous and heterozygous nature of NWR, gene duplication, and expression of numerous alternative splicing variants that were not completely collapsed by CD-HIT program (Haas et al., 2021). In this study, we selected relatively strict thresholds for BLAST searches to reduce “false positive” matches for assigning functions, which could have some effect on results of transcripts with no annotation (“false negatives”). This could also affect the number of NWR transcripts matching genes in other databases. The t_Boiv_ip contained 20,385 transcripts (N50 = 1,767 bp). All annotated DEGs at 48 h in this study were found in the total 10,674 fungal transcripts from infected NWR used to support the fungal genome assembly (Castell-Miller et al., 2016), and 94% matched a B. oryzae isolate from white rice (Condon et al., 2013). Our analyses were based on de novo transcriptome assemblies, and thus, the differential expression performed on transcripts of this pathosystem was independent of the draft genomes and had no effect on the evaluation and interpretation of downstream analyses.

Fungal infection triggered a distinct response of plant gene expression over time ( Figures 2 , 3 ; Table 2 ) compared to mock inoculation. The heatmap displayed distinct expression patterns in over half of the genes in response to fungal infection (upregulation) compared to mock inoculation (downregulation). This suggests a coordinated and specific response to the pathogen. Despite the less distinct expression of these genes at different time points, there was a greater stability at 48 h for both infected and mock groups. This could indicate a longer-lasting change in gene expression patterns for adaptation to environmental and/or biotic stress. In line with this, at least two principal components (PC1 = 32.2% and PC2 = 14.5%), which explained the variability in gene expression, were strongly associated with the treatments and collection time, respectively ( Supplementary Figure 4 ).

The number of NWR DEGs between fungal-infected and fungal-free plants increased by more than 2.6-fold over time ( Table 2 ). Although there were no significant GO enrichments in this set, many of the upregulated DEGs were associated with defense against biotic threats. The heightened activation of specific NWR genes in reaction to the existence of the fungus holds biological significance, as they possess functional annotations related to protection against biotic stresses ( Supplementary Table 13 ). These genes included membrane and cytoplasmic-localized receptors, downstream signaling events, elicitation of transcription factors, and the synthesis of various defense proteins, phytohormones, and antimicrobial secondary metabolites that are known to play a role in defense against pathogens (Wu et al., 2011; De Vleesschauwer et al., 2013; Camejo et al., 2016; Boutrot and Zipfel, 2017; Aldon et al., 2018; Chen et al., 2021). These events occur within a short time frame of 48 h, and their exact sequence and timing are not fully understood. However, this response is crucial for the plant’s defense against the fungus and highlights the complex and rapid nature of the plant’s immune system. This study has resulted in the initial framework for unraveling the interaction between NWR and B. oryzae ( Supplementary Figure 8 ).

Heatmaps showed visual contrasts between the transcriptional profiles of the fungus grown in vitro and in planta, highlighting its unique strategies for survival and growth in these different environments ( Figure 2 ). Time had less impact on gene expression, except for fungal growth in planta, where gene expression was more stable at 48 h and could indicate opposition to NWR gene expression. In agreement, the PCA detected that the conditions in which the fungus grows played a significant role in the variation of gene expression, with at least one PC with a strong association with treatments (PC1 = 28.4%) and another with a strong association with collection time (PC2 = 25%). Fungal DEG number increased over 2.3-fold from 24 h to 48 h. GO enrichments depicted that association with the host increased from 24 h to 48 h indicating an escalation in the plant–pathogen interaction. This likely is indicative of an extraordinary transcriptional and translational reprogramming involved not only in ordinary fungal metabolism for nutrient assimilation for growth and survival, but also for plant cell wall degradation ( Supplementary Table 8 ). Additionally, some of the upregulated fungal genes, including effectors, were important factors in the variancePartition analysis ( Figure 4 ). The data indicated fungal reprogramming for growth and pathogenesis over time as the engagement with NWR cellular defense activation progressed.

Plants have developed complex innate immune defenses against pathogens, which try to overcome plant tissue barriers, and/or suppress or manipulate their defenses (Toruño et al., 2016; Han and Kahmann, 2019; Shao et al., 2021). B. oryzae is generally considered a necrotroph (Condon et al., 2013), but it has also been suggested to have a short biotrophic stage (De Vleesschauwer et al., 2013; Thuy et al., 2023).

A major trigger for plant defenses is the detection of fungal chitin, the main fungal cell wall component, and its oligosaccharides (Vaghela et al., 2022). NWR DEGs with putative homology to two cell-surface LysM receptors, the RLP CEBiP (Kaku et al., 2006), upregulated at both time points, and the RLK OsCERK1 (Shimizu et al., 2010), upregulated at 48 h, were found. The RLP is essential for perception and binding to chitin derivatives through the external central LysM1 domain inducing dimerization (Hayafune et al., 2014), while OsCERK1 is required for chitin-elicitor signaling after hetero-oligomerization with CEBiP (Shimizu et al., 2010). In rice, chitin perception starts a series of downstream activation events (Akamatsu et al., 2013) including production of ROS, diterpenoid phytoalexins, callose deposition, and expression of basal defense genes (Shimizu et al., 2010; Ao et al., 2014). Both LysM chitin-induced receptors are needed to activate induced systemic resistance in rice against B. oryzae (Takagi et al., 2022). Perception of fungal chitin also activates acidic, exo-, and basic endo-chitinases (Vaghela et al., 2022). NWR produced transcripts similar to the rice vacuolar chitinases CHI1_ORYSJ (DN5450_c0_g1) and CHI2_ORYSJ (DN83338_c0_g1) that could have comparative defense roles as in rice (Datta et al., 2001). Within the most upregulated putative NWR TFs that could be linked to defense against fungi and response to chitin was a putative basic helix–loop–helix (bHLH041) (DN25752_c0_g1) gene. In wheat, a similar gene was uncovered as a participant candidate in a QTL conferring resistance to Fusarium head blight (Dhokane et al., 2016). B. oryzae also differentially secreted a protein with LysM domains (PF01476.21) and signal peptide (DN72040_c0_g1; CM_2799_g; Castell-Miller et al., 2016) matching Bipolaris/Cochliobolus spp. genes of carbohydrate binding module family 50, able to bind chitin and derivatives. It also matches, albeit with low similarity, the intracellular hyphae protein 1 of Colletotrichum lindemuthianum, important for its interaction with bean, including maintenance of its biotrophic phase (Perfect et al., 2000), and to the virulence effector Ecp6, a chitin-binding lectin (Bolton et al., 2008) from C fulvum; the latter avoids elicitation of PTI defenses (de Jonge et al., 2010) by outcompeting with AtCERK1 ectodomain receptor binding (Sánchez-Vallet et al., 2013), thus contributing to virulence.

Fungal endo-1,4-β-xylanases and a putative acetylxylan esterase, capable of hydrolysis of plant cell wall xylan, were differently expressed. In Botrytis cinerea, some xylanases also act as effectors inducing HR-like, ROS production and electrolyte leakage (Shao et al., 2021). Recently, a B. oryzae xylanase was reported as an apoplastic effector eliciting defenses in a Korean rice variety (Lee et al., 2023). NWR highly expresses xylanase inhibitors like those from rice (e.g., XIP-1 gene), which have competitive inhibitory activity toward enzymatic degradative activity of fungal xylanases (Durand et al., 2005). Other B. oryzae DEGs were two apoplastic effectors, pectate lyases, and a rhamnogalacturonan acetylesterase able to degrade pectin, present in the extracellular matrix of plant cell walls releasing oligogalacturonides (De Lorenzo et al., 2011). Putative NWR WAK membrane-bound receptors, capable of binding pectin and derivatives, were also DEGs. WAKs can activate MAPKs and other signaling pathways in response to cell wall disturbance (Kohorn, 2015). WAK receptors are involved in resistance to the spot blotch pathogen B. sorokiniana in barley at the seedling stage (Ameen et al., 2020) and to the hemibiotrophic Exserohilum turcicum causing northern corn leaf blight (Yang et al., 2019). Transcription of WAK genes through chitin detection in association with the CEBiP receptor was reported (Delteil et al., 2016). B. oryzae peptidases were differentially expressed. They could have a potential role as effectors and facilitate host colonization (Franceschetti et al., 2017). An extracellular metalloproteinase also carried a fungalysin domain, which is present in some effectors (Shao et al., 2021) like the gene Cgfl from Collectotrichum graminicola that is highly transcribed during the conversion from the biotrophic to the necrotrophic phase of corn colonization (Vargas et al., 2012), and could cleave plant chitinases (Sanz-Martín et al., 2016). A few NWR cysteine and serine protease inhibitors were also activated. Many other B. oryzae cell wall-degrading enzymes were also DEGs and found in NWR and white rice studies (Castell-Miller et al., 2016; Lee et al., 2023; Singh et al., 2023).

Among the most upregulated DEG matching receptors were several associated with plant defenses, such as SERK2, RLPK Feronia, and EIX2. SERK2 interprets cues of complex signaling networks from other surface-localized proteins and precisely modulates cellular responses including MAPK, Ca²⁺, ROS, and defense gene expression including phytoalexin production (Ma et al., 2016). It is associated with quantitative disease resistance to Sclerotinia sclerotiorum in Arabidopsis (Derbyshire et al., 2019) and positively regulates defenses against the blast fungus Magnaporthe oryzae in rice (Hu et al., 2005). The RLPK Feronia is another multi-functional receptor monitoring cell wall integrity (Kohorn, 2015) and defenses against M. oryzae in response to Ca²⁺ deficiency (Luo et al., 2022), but in tomato, it could facilitate fungal pathogenicity through mediating the Rapid Alkalinization Factor (RAFL)-FER pathway (Masachis et al., 2016). Both receptors in our pathosystem had a strong effect on treatment in the variancePartition analysis. EIX2 receptor (DN12795_c0_g1) is involved in receptor-mediated endocytosis and induction of HR (Ron and Avni, 2004).

Upon perception of fungal-derived conserved molecules, plant cell ROS homeostasis is altered (Camejo et al., 2016; Kawasaki et al., 2017). RBHOs are calcium-dependent NADPH oxidases that generate superoxide (Camejo et al., 2016) and participate in plant defenses. In our study, three potential NWR RBOH genes were DEGs, including an RBOHB and two RBHOA, of which ZPchr0010g9242 has a Ca²⁺ binding domain that could be involved in regulation of proteins (Torres and Dangl, 2005). Other DEGs were several peroxidases that can participate in the fortification of cell walls, detoxification (Camejo et al., 2016), and regulation of hormone signaling pathways of other PRPs and phytoalexins (Almagro et al., 2009). Involvement of ROS in rice defenses against B. oryzae has been reported (Ashfaq et al., 2021; Thuy et al., 2023). In our study, some of these genes were found within plant–pathogen interaction and MAPK KEGG biological pathways, and top drivers in the variancePartition due to treatments ( Supplementary Table 13 ), suggesting a strong ROS response during B. oryzae attack. The association of the complex signaling networking of RBOH genes with Ca²⁺, CDPK, and protein phosphorylation has been established in crops under abiotic and biotic (e.g., PAMPs) stresses (Torres and Dangl, 2005; Camejo et al., 2016).

Many NWR DEGs were associated with Ca²⁺ signaling pathways including plasma membrane receptors/ion channels involved in Ca²⁺ movement into the cell (Park and Shin, 2022), and Ca²⁺ binding sensors can detect spatiotemporal changes in the Ca²⁺ cytosolic content and, in turn, activate TFs, kinases, ROS, PRPs, and phytohormones (Cheval et al., 2013; Zhang et al., 2014; Aldon et al., 2018). Other associated genes included a putative NWR transcription regulator belonging to the CBP60 family, conserved in defenses against pathogens (Kumari et al., 2022), as well as an MLO-like protein, linked to the modulation of pathogen defenses (broad-spectrum) and programmed cell death in rice leaves (Kim et al., 2002). Ca²⁺ signaling genes were part of the plant–pathogen interaction and plant MAPK signaling KEGG enriched pathways (WGCNA), and a driving force in the variancePartition analysis.

Other signal molecules, MAPKs, were also differently expressed in our RNA sequencing study. Some of these MAPKs (M3K17 and M2K5) could be involved in abscisic acid regulation (Li et al., 2017), especially M2K5 that participates in ABA rice defense against B. oryzae (De Vleesschauwer et al., 2010). Another NWR transcript (DN4590_c0_g1) had high similarity to MPK12_ORYSJ (also called BWMK1), a gene activated by diverse stimuli including the rice pathogen M. oryzae, wounding, and ROS (He et al., 1999; Cheong et al., 2003) and in turn phosphorylates a TF that could interact with several PRPs and play a part in resistance to pathogens (Cheong et al., 2003).

Phytohormones involved in signaling processes have complex interactive networking that allows elicitation of specific defenses (Checker et al., 2018). In Arabidopsis, the SA signaling pathway mostly induces defenses against biotrophic stages of pathogens (Glazebrook, 2005) while the balancing action between/among JA, ET, and ABA pathways appears activated against some necrotrophic pathogens (Glazebrook, 2005; De Vleesschauwer et al., 2014; Ghozlan et al., 2020). However, rice phytohormone signaling does not appear to support the binary Arabidopsis model or to be tightly linked to the pathogen lifestyle, and it can even have participation of non-canonical phytohormones (De Vleesschauwer et al., 2013).

SA is an important component of plant immunity (Glazebrook, 2005; Qi et al., 2018) involved in local and systemic defense responses. SA signaling is linked to activation of defense-responsive genes (De Vleesschauwer et al., 2014) like PRPs and secondary metabolites (Glazebrook, 2005). In rice and in Brachypodium distachyon, sodium salicylate applications contributed to resistance against Rhizoctonia solani (Kouzai et al., 2018), and plants deficient in the production of SA (NahG mutants) were more susceptible. R. solani is considered a necrotroph; thus, it was suggested that a short biotrophic stage could be the target of rice SA defenses (Kouzai et al., 2018). NWR SA production can also be an attempt to control a possible subtle biotrophic phase of B. oryzae (De Vleesschauwer et al., 2013; Thuy et al., 2023). In our study, a putative B. oryzae salicylate hydroxylase enzyme (e.g., 7642_g) was highly upregulated, and it could participate in the degradation of aromatic compounds like SA or derivatives (Qi et al., 2018), or might alter the plant redox homeostasis provided by SA and thus promote oxidative damage (Yang et al., 2004) to favor its necrotrophic lifestyle phase. Moreover, B. oryzae also expresses a putative secreted ISC1 gene (DN46467_c0_g1; 4985_g), similar to the unorthodox effector from V. dahliae, which, in cotton, is required for complete pathogenicity by manipulating the SA metabolic pathways and suppressing defenses (Liu et al., 2014). However, the effectiveness of rice SA against B. oryzae infection was downplayed in other studies (Songkumarn et al., 2019).

JA and derivatives have a central role in rice defense signaling against pathogens with diverse lifestyles (De Vleesschauwer et al., 2013) with the majority of studies supporting jasmonates as positive regulators of disease control caused by hemibiotrophic (De Vleesschauwer et al., 2014) and some predominantly necrotrophic fungi (Kumar et al., 2023). However, in rice, JA-related genes were not effective defenses against B. oryzae (Ahn et al., 2005; De Vleesschauwer et al., 2014), or they have similar expression patterns in resistant and susceptible rice (Songkumarn et al., 2019). Other potential suggested roles support amplification of defense genes in response to diverse stressors (De Vleesschauwer et al., 2013), interacting with specific TFs to redirect the assistance of systemic defenses toward pathogen-inflicted wounds (Lorenzo et al., 2003), and regulation of ROS homeostasis in rice to alleviate oxidative ROS damage (Kumar et al., 2023).

ET immunity signaling appears to be complex and pathogen-dependent (De Vleesschauwer et al., 2014). In rice, production of ET positively restricts M. oryzae and R. solani invasion (De Vleesschauwer et al., 2013), but it acts as a strong suppressor of defenses against B. oryzae especially in susceptible cultivars (De Vleesschauwer et al., 2010). Several NWR ET-responsive transcription factors (ERF) were upregulated DEGs that could activate PRPs (Abiri et al., 2017) and participate in the regulation of ethylene/jasmonate pathways (Lorenzo et al., 2003) and, thus, warrant further investigation in this pathosystem.

The ABA signaling pathway has been found to be important for defense against B. oryzae in rice (De Vleesschauwer et al., 2010). External ABA application on leaves restricted fungal colonization in mesophyll cells by negating signaling expression of plant ET, which is thought to be induced by the fungus and promotes susceptibility. The ABA signaling occurs in an OsMPK5-dependent manner and with the recruitment of a heterotrimeric alpha subunit of a G-protein (De Vleesschauwer et al., 2010). This outcome is independent of SA and JA signaling pathways or callose deposition. ABA also appears to tune the effect of ROS dispersal and quantity over time on epidermal and mesophyll rice cells upon B. oryzae infection (De Vleesschauwer et al., 2010). Some NWR DEGs associated with ABA biosynthesis and the ABA signaling pathway were activated in reaction to B. oryzae infection ( Supplementary Table 13 ). For example, an NWR transcript in connection with ABA defenses was the exocytosis pathway E70B1, a positive regulator of ABA that assists in resistance against M. oryzae through perception of chitin and potentially interaction with CERK1 (Hou et al., 2020). Other ABA-related DEGs were MAPKs and GTPases.

Further research is needed to fully understand the specific roles of DEGs annotated to phytohormones within this pathosystem.

Phytoalexins appear to play a major role in the B. oryzae–NWR interaction. NWR DEGs matched those for phytoalexin production in rice and Z. latifolia. Rice produces several antimicrobial compounds, most of which are labdane-related diterpenoids (Lu et al., 2018) including momilactones, oryzalexins, and phytocassanes (Piasecka et al., 2015; Schmelz et al., 2014; Valletta et al., 2023). Rice phytocassanes could play a role against pathogens such as Magnaporthe oryzae and Xanthomonas oryzae (Lu et al., 2018). The NWR PBGC contains mainly genes putatively involved in phytocassane synthesis with a possible gene duplication of a copalyl-diphosphate synthase, which appears unique for the NWR cluster. Within this PBGC was also a CYP76M5 that could aid in phytoalexin production (Wu et al., 2011) and a putative kaurene synthase of the syn-pathway, which, in rice, is involved in the generation of a precursor of oryzalexin S (Nemoto et al., 2004). Another NWR DEG had similarity to a rice momilactone A synthase. Momilactones could act as phytoalexins, function in allelopathy (Xu et al., 2012), and play a part in non-host disease resistance (Lu et al., 2018).

Within fungal transcripts were effector peptides that could contribute to the onset and development of disease (Toruño et al., 2016; Shao et al., 2021). The vast majority corresponded to peptides without annotation. Others had similarities to known virulence genes such as isocytrate lyase, ICL1, of the hemibiotroph Leptosphaeria maculans associated with pathogenesis to canola (Brassica napus) (Idnurm and Howlett, 2002) and to a NPS and a PKS with roles in growth and development, reproduction, response to oxidative stress, and pathogenicity toward plants and microorganisms (Turgeon and Bushley, 2010; Condon et al., 2013).

Our NWR and B. oryzae global expression findings are in agreement with other results in the early stages of interactions between monocots and Bipolaris spp. such as B. sorghicola–Sorghum bicolor L (Yazawa et al., 2013), B. zeicola–corn (Liu et al., 2015), and recently published transcriptomes of the rice–B. oryzae pathosystem (Marwein et al., 2022; Lee et al., 2023; Singh et al., 2023).

The use of dual transcriptome analyses proved to be successful in identifying NWR genes responsible for defending against B. oryzae, as well as those that the fungus could utilize to promote disease. These genes displayed varying levels of differential expression at both 24 h and 48 h after inoculation.

These findings serve as a valuable foundation for guiding future research objectives. FBS affects a majority of NWR genotypes due to the scarcity of resistance. While this disease can be managed with fungicides, breeding for resistance poses significant challenges due to the species’ cross-pollinated nature, its aquatic habitat, and the likely quantitative inheritance of this trait. The few cultivars with improved resistance (e.g., Itasca-C12) struggle to maintain resistance in the field as the pathogen population increases throughout the season. However, the study findings provide candidate defense genes and associated pathways that can be explored at additional time points and for allelic diversity at the population levels, which could lead to the identification of potential markers across a wide range of transcripts/genes for crop improvement. Indeed, the identification of co-expressed defense and pathogenicity proteins/effectors is key to a comprehensive understanding of the NWR and B. oryzae pathosystem.

This study is the foundation for gene discovery toward a better understanding of NWR innate immunity against B. oryzae and molecular mechanisms used by the fungus to cause disease during the early colonization process. The study uncovered the induction of active and timely transcriptional reprogramming of membrane-bound and cytoplasmic receptors, signaling mechanisms, and activation of known disease-associated defense genes and antimicrobial phytoalexins. The candidate genes provide new avenues of research for exploring gene expression and allele sequencing analyses in FBS-resistant and -susceptible genotypes to sort out their function. They also could help develop markers for selection of NWR genotypes with improved resistance to FBS. In addition, B. oryzae virulence genes could be targeted for CRISPR/Cas9-based mutagenesis or RNA interference for an enhanced understanding of pathogenesis and host responses.

The data presented in the study are deposited in the Data Repository for the University of Minnesota (DRUM), accession link: https://doi.org/10.13020/9kja-aj88.

CC-M: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. TJYK: Data curation, Formal analysis, Resources, Validation, Visualization, Writing – review & editing. AR: Writing – review & editing. DCS: Software, Formal analysis, Writing – review & editing. DAS: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – review & editing. JAK: Funding acquisition, Project administration, Resources, Supervision, Writing – review & editing.