Regulatory balance between ear rot resistance and grain yield and their breeding applications in maize and other crops

Graphical abstract

Highlights

• •Pathogenic fungi infection causes significant reduction in crop yield and deterioration in grain quality.
• •Integrative analyses by using genetic and multi-omics data enable the identification of valuable genic resource.
• •Extensive syntenic relationships are observed among resistance-related loci for maize ER, wheat FHB, and rice RBD.
• •The genetic loci regulating both resistance and yield provide targets for manipulating the growth-defense trade-off.

Abstract

Background

Fungi are prevalent pathogens that cause substantial yield losses of major crops. Ear rot (ER), which is primarily induced by Fusarium or Aspergillus species, poses a significant challenge to maize production worldwide. ER resistance is regulated by several small effect quantitative trait loci (QTLs). To date, only a few ER-related genes have been identified that impede molecular breeding efforts to breed ER-resistant maize varieties.

Aim of review

Our aim here is to explore the research progress and mine genic resources related to ER resistance, and to propose a regulatory model elucidating the ER-resistant mechanism in maize as well as a trade-off model illustrating how crops balance fungal resistance and grain yield.

Key Scientific Concepts of Review: This review presents a comprehensive bibliometric analysis of the research history and current trends in the genetic and molecular regulation underlying ER resistance in maize. Moreover, we analyzed and discovered the genic resources by identifying 162 environmentally stable loci (ESLs) from various independent forward genetics studies as well as 1391 conservatively differentially expressed genes (DEGs) that respond to Fusarium or Aspergillus infection through multi-omics data analysis. Additionally, this review discusses the syntenies found among maize ER, wheat Fusarium head blight (FHB), and rice Bakanae disease (RBD) resistance-related loci, along with the significant overlap between fungal resistance loci and reported yield-related loci, thus providing valuable insights into the regulatory mechanisms underlying the trade-offs between yield and defense in crops.

Introduction

Sessile plants frequently encounter numerous detrimental invading organisms during their life cycle, among which fungi are the dominant causative agents of plant diseases [1]. In regard to agricultural production, fungi are the primary cause of large yield losses in major global crops such as maize, rice, and wheat [2]. Maize holds the top position worldwide in terms of yield per unit, total production, and planting area, followed by rice and wheat (FAOSTAT 2021, https://www.fao.org/faostat; data updated December 23, 2022). Maize is mainly utilized for food, fodder, and biofuel. Over the past century, its annual production has increased more than eight-fold through genetic advancements and agronomic management improvements to meet escalating global population demands [3], [4]. However, achieving enhanced tolerance against fungi, along with other biotic and abiotic stressors, could potentially improve maize yield [3].

The maize ear is a reproductive organ, and numerous pests and diseases have detrimental effects on its development and growth, resulting in significant yield loss and a decreased grain quality. Among them, ear rot (ER) caused by various pathogenic fungi is a major disease, and previous studies have reported grain yield reductions of 10–30 % attributed to ER [5], [6], [7]. In China, ER is widespread, particularly in the Huang-Huai-Hai, Northeast China, and the Southwest China maize production regions, leading to substantial economic losses [8], [9]. Three main types of ER have been discovered (Fig. 1A): (1) Fusarium ear rot (FER) is caused by F. fujikuroi species complex (FFSC) with F. verticillioides (Sacc.) Nirenberg being the most common species found throughout the world [6]. (2) Gibberella ear rot (GER) is mainly caused by F. graminearum species complex (FGSC) in temperate regions, with F. graminearum sensu strictu Schwabe (teleomorph Gibberella zeae) being the main causal agent [8], [10]. (3) Aspergillus ear rot (AER), which is mainly caused by A. flavus, occurs predominantly under hot and dry conditions [11], [12]. However, because of global climate change, all of these three types of ER can occur individually or together in most maize planting areas [7]. According to data in the GlobalFungi database [13], the genera Aspergillus and Fusarium exhibit a ubiquitous distribution in nature, with significant overlap observed specifically in croplands (Fig. 1B).

Fig. 1.

The effects of Aspergillus and Fusarium infection in worldwide croplands and the research hotspots revealed by bibliometric analysis. (A) The effects of Aspergillus and Fusarium infection on maize, wheat and rice.(B) The worldwide distribution of Aspergillus and Fusarium infection in croplands. (C) The keyword cluster timing diagram illustrating the differences between the main research topics involved in the different stages from 2000 to 2023.

The spikes and grains of other cereal crops such as wheat and rice are frequently subjected to devastating effects caused by fungi during development and growth (Fig. 1A). For instance, Fusarium head blight (FHB), caused by Fusarium species, poses a global threat to wheat and other small-grain cereal production [14], [15], [16]. Rice Bakanae disease (RBD), which is induced by Fusarium species and transmitted through seeds, is becoming an increasingly significant concern in several rice-growing countries [17], [18].

To establish an ecological advantage over other organisms that share the same ecosystem and also to enhance environmental adaptability, most pathogenic fungi colonize host plants by producing mycotoxins under specific environmental conditions [19], [20]. For example, F. graminearum produces at least seven mycotoxins, with deoxynivalenol (DON) and zearalenone (ZEN) being the main types; F. verticillioides mainly produces three B types of fumonisins (FB1, FB2, and FB3) [6]; and A. flavus primarily produces aflatoxins B1 and B2 [21]. Most mycotoxins can induce various toxicological effects through the consumption of contaminated food, including reduced livestock productivity, immune deficiency in humans, and even cancer [22], [23]. Consequently, grains exceeding the recommended toxicity limits are rejected in the global crop trade, resulting in significant annual financial losses amounting to hundreds of millions of US dollars [24], [25]. Pathogenic fungi-induced ER can also infect maize seedlings via the seeds and subsequently cause stalk rot and ERs in the subsequent generations [26].

The identification of resistant germplasms and utilization of host resistance are the most environmentally safe and cost-effective strategies for controlling maize ER. Although complete resistance has not yet been achieved, past studies have reported the existence of ER-resistant germplasm in maize and its ancestral teosinte populations [27], [28], [29], [30], [31]. Multiple different approaches, such as quantitative trait loci (QTL) mapping and genome-wide association analysis (GWAS), have been employed in order to identify genetic loci and candidate genes that regulate ER resistance and mycotoxin contamination in germplasms with varying levels of susceptibility [27], [28], [29], [32]. However, genetic analysis has revealed that all types of ER resistances are controlled by a multitude of small-effect QTLs which are strongly influenced by environmental factors [12], [33], [34], [35], [36].

To date, only a few genes have been cloned and validated in regard to regulating ER resistance [37]. Comprehensive multi-omics investigations, including transcriptomics, proteomics, and metabolomics, can be performed to identify potential candidate genes and elucidate the molecular mechanisms underlying maize resistance to fungi. This review aims to comprehensively summarize the research progress made in diverse genetics, multi-omics, and functional genomics towards elucidating the genetic and molecular mechanisms underlying ER resistance based on integrative biological analyses (Material S1). We have analyzed the extent of convergent evolution among grass species at a genome-wide level concerning fungi resistance loci and evaluated how this knowledge can facilitate the identification of novel resistance genes in crops. Importantly, we have investigated correlations between the genetic loci of ER resistance and grain yield and refined potential loci involved in maintaining a delicate balance between crop yield and fungal defense, thereby providing valuable genetic and genic resources for enhancing disease resistance and improving crop yield simultaneously.

Research history and current trends on the genetic and molecular regulations of ER resistance in maize

Since the 1960 s, researchers have consistently observed stable differences in ER severity among maize varieties (including both inbred lines and hybrids) across various environments, indicating that ER resistance is a heritable trait regulated by genetic control [38], [39]. Subsequent studies revealed that ER resistance is governed by multiple different genes whose effects are influenced by environmental conditions [40], [41], [42]. Inheritance investigations have demonstrated that all three types of ER (FER, AER, and GER) resistance are controlled by additive gene effects as well as non-additive effects, such as dominance and epistasis [40], [43], [44]. Given the prevalence of additive gene action in regulating ER resistance in maize, it is feasible to enhance the ER resistance by pyramiding favorable gene combinations.

However, past studies on QTL mapping for ER resistance have primarily been conducted since 2000. Pérez-Brito et al. (2001) initially reported QTLs associated with genetic resistance to F. moniliforme (synonym F. verticillioides), with only three overlapping QTLs between the two investigated populations [45]. The first report of AER-related QTLs was conducted by Paul et al. (2003), who found that aflatoxin accumulation was regulated by multiple small effects and unstable QTLs [30]. Ali et al. (2005) reported 11 and 18 GER-associated QTLs following silk and kernel inoculation in four different environments, respectively; however, only two and one common QTLs, respectively, were detected among these tests, suggesting a strong environmental influence on these traits [46].

In order to further elucidate the genetic and molecular mechanisms underlying ER resistance, we retrieved 226 published articles from the Science Citation Index Expanded (SCIE) database of Web of Science (Table S1). Substantial progress has been made in unraveling the genetic basis of ER resistance in maize, as evidenced by the high-frequency and centrality analyses of key terms. Notably, the quantitative trait loci, Fusarium verticillioides, ear rot, Aspergillus flavus, resistance, contamination, Fusarium graminearum, and mycotoxins emerged as the top ten keywords (Fig. S1A, Table S2), highlighting the importance of identifying the QTLs for resistance against fungi and their accompanying mycotoxins.

The keyword cluster timing diagram aids in distinguishing differences between the main research areas at various stages. Nine clusters labeled with the symbol # were obtained from 2000 to 2023 (Fig. 1C) based on the methodologies described by Xing et al. (2024) [47]. Representative keywords with high co-term frequencies in each cluster indicated that quantitative trait loci (in #0), genes (in #2 and #5), and mycotoxins (in #1, #3, and #4) were the primary research areas (Fig. 1C). The 25 keywords with the strongest citation bursts (Fig. S1B), illustrated the emergence and replacement of research hotspots over the last 20 years. ERs can be infected through the silk or kernel in the field [48], [49], [50], and the appearance of the keyword “kernel infection” in the year 2000 along with its burst in 2012 indicates that researchers have more interest in studying kernel resistance (Fig. 1C, S1B). Artificial inoculation methods that imitate silk or kernel infections have been adopted in most studies [7], [51]. However, several milliliters of conidial suspension are generally used for fungal injection [52], [53], which results in kernel infection. Therefore, we differentiated among the three types of ERs, but not the inoculation methods, in the following analysis.

GWAS have been employed for the genetic dissection of ER resistance since 2013 [54], with notable citation bursts observed in 2017 (Fig. 1C, S1B). Notably, there is a recent surge in citations of genomic selection (GS) in 2021 (Fig. 1C, S1B), indicating that the rapid accumulation of information on the genetic basis of ER resistance has motivated researchers to enhance ER resistance using genome-wide technologies.

Genetic structure of maize ER resistance revealed by QTL mapping and GWAS

Considering that ER resistance is regulated by small-effect and environmentally unstable QTLs [27], [30], [45], refining environmentally stable loci (ESLs) that can be detected in multiple independent studies would be able to facilitate the identification of authentic QTLs and enable the application of marker-assisted selection (MAS) or GS in maize breeding programs. In order to achieve this goal, published articles focusing on QTL or quantitative trait nucleotide (QTN) identification were retrieved to identify the ER-resistance ESLs in maize (Table S1). Four articles in Chinese focusing on the genetic analysis of ER resistance in high-incidence ER areas of China, particularly in the Huang-Huai-Hai maize production region, were also included [55], [56], [57], [58]. A total of 36 and 25 published papers on QTL mapping and GWAS for ER resistance, respectively, were retrieved for further analysis (Table S1). All chromosomal positional information of the original QTLs and QTNs was normalized and projected onto the reference B73 genome (Zm-B73-REFERENCE-NAM-5.0), according to the methods of Dong et al. (2023) [59]. Consequently, a total of 335 QTLs and 3662 QTNs were normalized (Fig. S2, Tables S3-S4).

For AER resistance, a total of 100 normalized QTLs were identified across ten maize chromosomes (Chrs), with Chr 2 and Chr 5 harboring the highest number of QTLs (16) (Fig. S2A, Table S3). Consequently, a set of 12 QTL hotspots that were consistently mapped by three or more independent studies was refined. These AER QTL hotspots were found on Chr 1, 2, 3, 4, 5, and 7, with intervals spanning from 0.79 to 4.49 Mb (Fig. 2A, Table S5). In addition, a total of 352 QTNs were identified across all ten chromosomes. Among them, Chr 1 and Chr 7 had the highest (55) and lowest (20) number of QTNs, respectively (Fig. S2B, Table S4). We defined a 0.5-Mb sliding window containing three or more normalized QTNs as a QTN hotspot. Consequently, a total of five QTN hotspots were refined with the intervals ranging from 0.50 to 0.75 Mb (Fig. 2A, Table S6).

Fig. 2.

Summary of the genetic hotspots for ER resistances. (A) Chromosomal distributions of the five types of genetic hotspots for ER resistances. The first layer represents the ten maize chromosomes, while the second to sixth layers depict the chromosomal distributions of the QTN hotspots of FER, QTL hotspots of FER, QTN hotspots of AER, QTL hotspots of AER and QTL hotspots of GER, respectively. (B) Chromosomal distributions of the nine mutually verified QTL/QTN hotspots for FER resistance by both QTL mapping and GWAS. (C) Chromosomal distributions of the 11 pleiotropic genetic hotspots for resistance to at least two types of ERs.

For GER resistance, a total of 91 QTLs were identified across all ten chromosomes, with Chr 1 and Chr 10 harboring the highest (18) and lowest (3) number of QTLs, respectively (Fig. S2A, Table S3). Nine QTL hotspots with intervals from 0.26 to 9.30 Mb were identified after removing two hotspots with too large confidence intervals (41.04 Mb and 45.95 Mb, respectively). Notably, most QTL hotspots were found on Chr 3 and Chr 5 (Fig. 2A, Table S5). No GER QTN hotspots were identified primarily because only seven QTNs were originally retrieved (Fig. S2A, Table S4).

For FER resistance, a total of 144 normalized QTLs across maize ten chromosomes were identified, with Chr 4 and Chr 7 exhibiting the highest (23) and lowest (6) number of QTLs (Fig. S2A, Table S3). Based on these data, a total of 37 QTL hotspots were detected among the seven chromosomes, excluding chromosomes 7, 9, and 10. However, one QTL hotspot with an unusually large confidence interval of 69.66 Mb and three QTL hotspots with extremely small intervals ranging from 89 to 1786 bp were excluded from further analysis (Fig. 2A, Table S5). Additionally, a collection of 3303 QTNs was obtained across all ten chromosomes, with Chr 2 and Chr 10 exhibiting the highest (476) and lowest (223) number of QTNs, respectively (Fig. S2B, Table S4). Subsequently, a set of 103 FER-related QTN hotspots across all ten chromosomes were detected at intervals ranging from 0.52 to 1.96 Mb (Fig. 2A, Table S6).

In total, 162 ER-resistant ESLs, including 54 QTL and 108 QTN hotspots with the total of 148 non-overlapping regions, were refined. Every specific QTL hotspot was supported by 3–6 independent studies (Table S5), whereas a single QTN hotspot encompassed 3–31 QTNs (Table S6). Additionally, initial validation has been conducted on 18 molecular markers residing in 17 ESLs to evaluate their association with ER resistance (Table S5-S6), we propose that these markers exhibit promising potential in assisting molecular breeding.

QTL mapping and GWAS are two complementary technologies that have been used to elucidate the genetic architecture underlying complex trait variations in plants [60]. Genetic loci identified using both methods were deemed reliable ESLs. We identified nine mutually verified QTL/QTN hotspots for FER resistance on Chr 2, 3, and 6 (Fig. 2B). However, the absence of mutually verified QTL/QTN hotspots for both AER and GER resistance may be attributed to the limited availability of data. We propose that the pleiotropic genetic hotspots associated with multiple types of ER resistance may harbor pivotal genes responsible for broad-spectrum ER resistance. Moreover, previous studies have identified pleiotropic QTLs that confer resistance to FER and AER [61]. By integrating all QTL and QTN hotspot data, 11 pleiotropic genetic hotspots regulating at least two types of ER resistance were identified on Chr 1, 2, 3, 5, and 7 of maize genome (Fig. 2C). Notably, five overlapping pleiotropic genetic hotspots were found to be enriched in the genomic region of 189.15–233.34 Mb on Chr 3 (Fig. 2C), suggesting a pivotal role for this region in conferring broad-spectrum resistance against ER in maize.

Genic resources for ER resistance by combining genetics and multi-omics analyses

Based on the annotation of the B73 genome, the confidence intervals harboring the above 162 ER-resistant ESLs encompassed a total of 7228 candidate genes (Table S7). To identify the ER-responsive genes, a combined analysis of eight sets of published omics data was conducted.

Among the five transcriptomic datasets for FER [62], [63], [64], [65], [66], the number of differentially expressed genes (DEGs) shared by five, four, three, and two studies were 3, 20, 174, and 7057, respectively (Fig. 3A). Furthermore, a comparison of the 7057 DEGs responsive to FER with the 1113 DEGs responsive to GER [67], 185 DEGs responsive to AER [68], and 1782 GER-responsive differentially expressed proteins (DEPs) identified through proteomics research [69] revealed a total of 1391 cross-validated genes, with at least twice being detected (Fig. 3B, Table S8). In detail, 85 % (946 vs. 1113) of the 1113 GER-DEGs and 44 % (81 vs. 185) of the 185 AER-DEGs could be cross-validated using other datasets (Fig. 3B), indicating that the convergent regulatory pathways can be activated in response to diverse pathogenic infections causing ER in maize.

Fig. 3.

Candidate genes responsive to Fusarium and Aspergillus infection. (A) Identification of the inducible genes in response to Fusarium verticillioides infection through analysis of five transcriptomic datasets. Numbers of the DEGs shared by five, four, three and two studies are 3, 20, 174, and 7057, respectively. (B) Comparison of the 7057 DEGs responsive to FER with the 1113 DEGs responsive to GER, the 185 DEGs responsive to AER, and the 1782 GER-responsive DEPs that retrieved from various research studies. In total, a set of cross-validated genes consisting of 1391 candidates detected in at least two datasets were identified. (C) Circos plot depicting the chromosomal distributions of the genes derived from the different levels of analysis. The first layer represents the ten maize chromosomes, followed by the second layer illustrating chromosomal positions of 7228 candidate genes identified in 162 ESLs. The third layer displays chromosomal locations of 1391 cross-validated DEGs obtained from various multi-omics datasets, and the fourth layer exhibits chromosomal distributions of 297 DEGs located within the aforementioned 162 ESLs. The fifth layer highlights chromosomal positions of 65 genes exhibiting enrichment in crucial pathways and potentially contributing to resistance against ER. Lastly, the sixth layer signifies chromosomal locations of 12 known genes correlated with ER resistance. (D) Summary of the candidate genes derived from the different levels of analysis. Levels 1 to 4 (L1 to L4) represent the constituents of 7228 candidate genes in 162 ESLs, the 1391 ER-responsive genes, the 297 ER-responsive genes within ESLs, and the 65 GO/KEGG enriched genes derived from 297 ER-responsive genes within ESLs, respectively. The bar graph “6931 genes” refers to non-DEGs located within the 162 ESLs, “1094 genes” represents ER-related DEGs found outside these ESLs, while “232 genes” indicates non-enriched DEGs by GO/KEGG analysis within the 162 ESLs.

Gene Ontology (GO) analysis revealed that the 1391 DEGs were mainly enriched in resistance-related GO terms, such as oxidation–reduction process, which was ranked as the most prominent, defense response, response to biotic stimulus, oxylipin metabolic process, organic acid metabolic process, and glutathione metabolic process (Fig. S3, Table S9). In addition, genes involved in carboxylic acid and cellular modified amino acid metabolic processes were also enriched (Fig. S3, Table S9), suggesting that the genes involved in plant development and growth also respond to ER fungal infections. Similarly, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed significant enrichment of resistance-related pathways such as glutathione metabolism, phenylpropanoid biosynthesis, flavonoid biosynthesis, and alpha-linolenic acid metabolism, as well as plant development and growth-related pathways such as carbon metabolism, glycolysis/ gluconeogenesis, carbon fixation in photosynthetic organisms, and also the citrate cycle (Fig. S4, Table S10). Notably, phenylpropanoid biosynthesis and carbon metabolism exhibited conserved enrichment following infection with all three types of fungi (AER, FER, and GER) [62], [69].

Further comparison of the above 1391 DEGs with the 7228 candidate genes in the 162 ER-resistant ESLs identified a total of 297 overlapping DEGs, which were associated with 96 ESLs (Fig. 3C, Table S11). Among them, 31 (32.29 %) ESLs contained only a single DEG and 93 (96.88 %) ESLs encompassed no more than ten DEGs (Table S11), suggesting that the limited number of candidate genes located in specific ER-resistant ESLs raises the possibility of being able to rapidly validate the causal genes of ESLs and their functions by manipulating them through gene editing or overexpression methods. Furthermore, the 297 DEGs displayed eight distinct expression patterns. Among them, 10 DEGs (3 %, Class I) were specifically up-regulated in resistant lines, while 189 DEGs (64 %, Class II) were up-regulated in both resistant and susceptible lines (Table S11). Moreover, we observed that 13 DEGs (4 %, Class VI) showed specific down-regulation exclusively in the resistant lines, whereas 31 DEGs (10 %, Class VII) were commonly down-regulated in both resistant and susceptible lines (Table S11). Following this filtration step, a higher proportion of the 297 genes exhibit significant enrichment in resistance-related pathways (Fig. S5). The expression of resistance genes, such as pathogenesis-related (PR) genes, are typically induced in resistance germplasms [70], while susceptibility genes tend to be inhibited [71], [72]. Some PR genes, namely Zm00001eb150050 (PR3), Zm00001eb032600 (PR5), and Zm00001eb013990 (PR10), were concurrently identified and observed to be up-regulated in both resistant and susceptible lines (Table S11). These findings imply that these PR genes may serve crucial roles in response to pathogen challenge. Furthermore, we identified 65 genes that were enriched in crucial pathways and potentially contributed to ER resistance (Fig. 3C-D, Table S12). Among them, 12, 11, 8, 7, and 5 genes were enriched in the phenylpropanoid/flavonoid metabolism, carbon metabolism, amino acid biosynthesis, plant-pathogen interaction, and glutathione metabolism pathways, respectively (Table S12). Notably, three of the 12 known genes correlated with ER resistance were found to be DEGs (Fig. 3C, Table S8), thereby reinforcing the significance of integrating multi-omics data for the discovery of valuable genic resources.

The conserved genetic basis underlying phenotypic convergence of fungi resistance among maize, wheat, and rice

FHB and RBD are two fungal diseases mainly caused by Fusarium in wheat spikes and rice seeds, respectively, resulting in worldwide crop production loss [16], [18]. Given the relatively close evolutionary relationships between maize, wheat, and rice [73], it is intriguing to investigate whether the phenotypes of fungal resistance among these three species share similar genetic and molecular bases.

In order to achieve this goal, 579 FHB-resistant loci in wheat (Fig. 4A, Table S13) and 31 RBD-resistant loci in rice (Fig. 4B, Table S14) were identified. Of the 162 ER-resistant ESLs, 169, 163, and 164 syntenic blocks containing 3090, 2998, and 3064 orthologous genes in the wheat A, B, and D genomes, respectively, were identified (Fig. 4C, Table S15). Interestingly, 45 (26.63 %), 51 (31.29 %), and 24 (14.63 %) of these syntenic blocks, encompassing a total of 356, 227, and 104 orthologous genes in the wheat A, B, and D genomes, respectively, reside in FHB-related QTL/QTN regions (Fig. 4C, Table S16). Four major QTLs (Fhb1, Fhb2, Fhb4, and Fhb5) conferring resistance to FHB in wheat have been identified, with successful cloning of the Fhb1 gene [14], [15], [74], [75], [76]. No syntenic blocks were identified between the wheat Fhb1 locus and maize ER HSs. However, loci of Fhb2 and Fhb4 are collinear with ER-resistance ESLs Chr 4:239.49–240.22 Mb and Chr 9:159.32–161.22 Mb, respectively (Table S16), suggesting that these syntenic blocks may play similar roles in regulating ER and FHB resistance between maize and wheat.

Fig. 4.

Conservation of the fungi resistance loci among maize, wheat and rice genomes. (A) The genetic loci for FHB resistance in wheat. A total of 579 FHB-resistance loci in wheat were collected. (B) The genetic loci for RBD resistance in rice. A total 31 RBD-resistance loci in rice were collected. (C) Syntenic relationships among the ER-resistance ESLs with wheat and rice genomes. A total of 162 ER-resistance ESLs (148 non-overlapping regions), with 169, 163, and 164 syntenic blocks in the wheat A, B, and D genomes respectively, as well as 153 syntenic blocks in the rice genome were found. Among these syntenic blocks, 45, 51, and 24 were located within FHB-related QTLs/QTNs regions in wheat A, B, and D genomes while an additional 12 syntenic blocks resided within rice RBD QTLs. An example illustrating an ER-resistance ESL on chromosome 3 along with its corresponding syntenic blocks in both wheat and rice was presented under the circus plot.

In total, 153 syntenic blocks of 162 ER-resistant ESLs containing 3524 orthologous genes were identified in the rice genome (Fig. 4C, Table S17). Among these, 12 (7.8 %) RBD QTLs that harbor 217 orthologous genes were identified (Fig. 4C, Table S18). Several major QTLs for RBD resistance on rice Chr 1, including qBK1^Z (1.43–2.16 Mb), qBK1^WD (13.54–15.13 Mb), and qFfR1 (22.56–24.10 Mb) have been detected [77], [78], [79]. Interestingly, both AER and FER hotspots on maize Chr 3 were found to be located in the syntenic regions of these RBD resistance QTLs (Tables S5, S6, S17), suggesting that the candidate gene(s) in the syntenic regions may have evolved convergently in order to exert analogous regulatory roles in fungal resistance between maize and rice.

In plants, the conservation of orthologous genes is one of the primary molecular mechanisms underlying phenotypic convergence [80], [81]. Although the morphologies of maize, wheat, and rice exhibit substantial divergence from their common ancestors [73], a number of adaptation-associated traits share common molecular mechanisms [82]. These crops frequently encounter comparable external stimuli during cultivation, such as fungal infections by Fusarium species [16], [17], [18]. Our findings suggest that there may be a shared genetic basis for phenotypic convergence in fungal resistance among the three species, providing valuable references and guidance for novel resistance gene discovery in different crops. For example, seven sets of conserved syntenic blocks (Chr 2: 9.02–9.17 Mb, Chr 3: 6.61–7.19 Mb, Chr 3: 8.57–10.68 Mb, Chr 3: 14.89–15.01 Mb, Chr 3: 219.11–219.39 Mb, Chr 3: 226.46–226.47 Mb, and Chr 8: 20.69–20.93 Mb) harboring fungal resistance QTLs/QTNs were identified across the three species (Fig. 4C, Tables S15, S17). It would be worthwhile to investigate whether the functional conservation of orthologous genes within these syntenic blocks simultaneously contributes to ER, FHB, and RBD resistance.

In conjunction with omics data, 42 DEGs in maize, along with 35 and 12 orthologous genes residing in the confidence intervals of the wheat and rice QTL and QTN regions, respectively, were identified (Table S19). Importantly, among them, three maize DEGs that encode RING-type E3 ubiquitin transferase (Zm00001eb123630), heat shock protein (HSP, Zm00001eb121900), and glycosyltransferase (Zm00001eb122610) were simultaneously identified as orthologous genes in rice and wheat (Table S19), and have also been reported to be associated with biotic stresses [83], [84], [85]. It is therefore plausible that these orthologous genes exhibit conserved functions in fungal resistance across crop species.

Molecular basis and regulatory network underlying ER resistance in maize

Summary of the functionally validated genes with ER resistance

To date, the molecular regulatory network underlying maize ER resistance remains largely unknown, and only 12 genes have been cloned and functionally validated for ER resistance (Fig. 3C). Among these, mitogen-activated protein kinase 3 (ZmMAPK3) has been identified as a candidate gene for the FER QTL qFER1.03, and its RNAi lines show a significant decrease in FER resistance [65]. ZmSIZ1a and ZmSIZ1b encode SUMO E3 ligases that respond to various stresses [86]. Their double mutants show severely impaired resistance to FER and they are redundant, but indispensable for combating FER. Transcriptome analysis using the double mutant revealed that DEGs were enriched in plant disease resistance-related pathways, including MAPK signaling, plant phytohormone signal transduction, and flavonoid metabolite synthesis [87].

Lipoxygenases (LOXs) have been shown to participate in the formation of oxylipins, which are signaling compounds involved in host-fungus crosstalk via polyunsaturated fatty acid oxidation pathways [88], [89]. Enhanced expression of LOX genes following F. verticillioides infection in maize has been observed [90], and mutant analysis has further revealed that ZmLOX4, ZmLOX5, and ZmLOX12 positively but ZmLOX3 negatively regulate ER resistance via jasmonic acid (JA)-dependent pathways [91], [92], [93], [94].

1-aminocyclopropane-1-carboxylate (ACC) synthases catalyze the conversion of S-adenosyl-L-methionine (SAM) to ACC during Ethylene (ET) biosynthesis [95]. Knockouts of ZmACS2 and ZmACS6 abolished ET accumulation and reduced F. verticillioides colonization in maize seeds [96]. A gene (Zm00001eb193360) associated with ET synthesis was found to be down-regulated simultaneously in both resistant and susceptible lines upon pathogen challenge (Table S11). Therefore, it is intriguing to investigate the potential roles of this gene in conferring ER resistance through further experiments. The causal gene of stalk rot resistance QTL qRfg2 was identified as ZmAuxRP1, and phenotyping of both near-isogenic lines (NILs) and transgenic lines showed that this auxin-regulated protein simultaneously mediates maize resistance to both stalk rot and ER [97].

ZmFER1, a homolog of the wheat FHB resistance gene TaHRC, encodes a histidine-rich calcium-binding protein [14], [15], and negatively regulates resistance to FER [98]. Interestingly, a deletion spanning the start codon of TaHRC and a 6-bp deletion in the ZmFER1 histidine-rich coding region in natural populations are closely associated with FHB and FER resistance, respectively [15], [98], which provides an example of the convergent evolution of Fusarium resistance at the molecular level between maize and wheat. TaHRC regulates cation exchanger-interacting protein 4 (TaCAXIP4), which maintains Ca²⁺-homeostasis [99]. These functionally validated ER resistance genes are important gene resources and key regulators in the construction of the molecular regulatory network underlying ER resistance in maize (Fig. 5).

Fig. 5.

A proposed molecular network regulating ER resistance, and the gene resources optimizing maize resistance. Upon perception of PAMPs by plant cell-surface-localized PRRs and formation of receptor complexes with co-factors such as RLCKs (PTI pathway) or recognition of effectors by intracellular NLRs (ETI pathway), the signals converge at various downstream outputs, including MAPK cascades, calcium influx, ROS bursting, transcriptional reprogramming, and phytohormone signaling. Ultimately, the synthesis of defense-related components, primarily phenylpropanoid compounds, or the induction of programmed cell death is triggered. Candidate genes derived only from the DEGs are labeled in black, and DEGs located in the ESLs are labeled in red. Solid arrows indicate the direct effects, and dashed arrows indicate the indirect effects. C2, C2 domain Ca²⁺ sensor; CPK, calcium-dependent protein kinase; DEGs, differentially expressed genes; DRGs, defense related genes; ESLs, environmentally stable loci; ETI, effector-triggered immunity; GSH, glutathione; L, laccase; NLRs, nucleotide-binding domain, LRR containing receptors; MAPK, mitogen-activated protein kinase; MAPKK, mitogen-activated protein kinase kinase; MAPKKK, mitogen-activated protein kinase kinase kinase; P, peroxidase; PAMPs, pathogen-associated molecular patterns; PMI, pectin methylesterase inhibitor; PRGs, phytohormone related genes; PRRs, pattern recognition receptors; PTI, pattern-triggered immunity; RBOH, respiratory burst oxidase homologue; RLCKs, receptor-like cytoplasmic kinases; ROS, reactive oxygen species; SOD, superoxide dismutase; TFs, transcription factors.

The ZmCCT gene, which contains a CCT domain, has been reported to confer resistance to F. graminearum and is identified as the causal gene for the stalk rot resistance QTL qRfg1 [100]. Initially, ZmCCT was discovered as a mediator of photoperiod sensitivity, and its responsiveness to photoperiodic conditions is attenuated by an insertion of a CACTA-like transposable element (TE1) in the promoter region, thereby accelerating maize spread to long day regions [101], [102]. However, the non-TE1 allele of ZmCCT exhibits disease resistance, whereas the TE1 allele is susceptible to F. graminearum infection [100]. Multi-omics analysis reveals that ZmCCT potentially functions through salicylic acid, auxin, phenylpropanoid biosynthesis, and phenylalanine metabolism pathways [103]. Notably, a haplotype exhibiting strong performance in both flowering time and F. graminearum resistance holds significant potential for future maize breeding programs [104].

Er-resistance genes and pathways revealed by multi-omics investigations

The integration of multi-omics data will facilitate the understanding of maize responses to fungal infection and the identification of candidate genes for ER resistance. Among the three main types of ERs (AER, FER, and GER), extensive multi-omics investigations have mainly focused on the response of maize to FER, and the resistant line CO441 and susceptible line CO354 have been repeatedly used. Using microarray gene expression technology, Lanubile et al. (2010) found similar FER responses at 48 h post-inoculation (hpi) in both genotypes [105]. Notably, defense-related genes encoding PR proteins, HSPs, cytochrome P450, peroxidase, and detoxification response proteins were found to be significantly upregulated, whereas cell metabolism- and development-related genes were found to be predominantly downregulated. In addition, defense-related genes exhibited high transcriptional levels even before infection occurred in CO441, suggesting their potential roles in providing basal defense against fungal pathogens. Lanubile et al. (2014) found that the oxidative burst and auxin pathway activation occur starting at 48 hpi onward, and that at 72 hpi, genes encoding receptor-like kinases (RLKs), protein kinases, calcium homeostasis regulators, WRKY, NAC, MYB transcription factors (TFs), and JA/ET signaling members are commonly activated [63]. Notably, FER specifically induced genes in CO441 were mainly related to secondary metabolism, including shikimate, lignin, flavonoid, and terpenoid biosynthesis. However, activation of defense- and signal transduction-related genes was exclusively observed in the susceptible line L4637 at 72 hpi following F. verticillioides infection [106].

Investigation of the early response to F. verticillioides infection showed that the DEGs occurred as early as 5 min after pathogen treatment, and plant phytohormone signal transduction and phenylpropanoid biosynthesis emerged as the top two pathways observed throughout the entire 6-hour infection process [66]. The dynamic transcriptional response to F. verticillioides infection in the resistant line BT-1 revealed that DEGs were primarily enriched in sucrose metabolism, protein processing in the endoplasmic reticulum, and amino sugar and nucleotide sugar metabolic pathways at an early stage (24 hpi). However, antibiotic and terpenoid biosynthetic processes, such as phenylpropanoid biosynthesis, are preferentially induced during the late stage (10 days post-inoculation (dpi)), indicating the potential conversion of secondary metabolites into antifungal compounds [64]. Using a bulk-segregant RNA-seq approach, Cao et al. (2022) observed the downregulation of genes involved in sink metabolic processes and the upregulation of genes associated with secondary metabolism and compounds/processes, particularly those related to cell wall biosynthesis/rearrangement and flavonoid biosynthesis, in resistant inbred bulks at 10 dpi following F. verticillioides infection [62]. These findings suggest a growth-defense trade-off following prolonged infection with F. verticillioides.

Inoculation of F. graminearum at 24 and 48 hpi between the resistant line CO441 and the susceptible line B73 revealed that early-induced genes were involved in secondary metabolism and cell wall modification/biosynthesis, whereas late-induced genes were mainly involved in auxin, ET, JA, salicylic acid (SA), and abscisic acid (ABA) metabolism, MAPK signaling, and ERF, WRKY, and MYB TFs [67]. Proteomic profiling at 48 hpi demonstrated an increased abundance of defense response proteins and also cell wall modification/ biosynthesis proteins in these two inbred lines [69].

DEGs between the AER-resistant line Mp715 and the susceptible line B73 were investigated after 2 and 3 weeks of A. flavus infection [68], and pathways of carbohydrate metabolism, sugar sensing, programmed cell death, flavonoids, phytoalexins, and lignin biosynthesis were found to be significantly enriched. Notably, these types of DEGs are largely consistent with the enriched DEGs identified here, as listed in Table S8.

Overall, maize resistance to ER is a multifaceted and dynamic process that remains poorly understood and it is characterized by conflicting findings reported in different studies. However, consistent results indicate that defense-related genes exhibit relatively high expression levels in resistant lines prior to infection, suggesting that basal and constitutive defence levels can confer a significant advantage to the host against threats from F. verticillioides, F. graminearum and A. flavus [69], [105], [107]. Representative genes encoding Ca²⁺ sensors, R proteins, WRKYs, and cell-wall modification/biosynthesis factors were selected to construct the molecular regulatory network underlying ER resistance in maize (Fig. 5).

Comparison of ER-resistance genes with components of plant immune systems

Plants have developed two layers of innate immune system that recognize and counteract diverse pathogen attacks. The initial layer is known as pattern-triggered immunity (PTI), which is activated by pathogen-associated molecular patterns (PAMPs) and it is perceived by plant cell surface-localized pattern recognition receptors (PRRs). PRRs are usually plasma membrane-bound RLKs or receptor-like proteins (RLPs) with extracellular domains that allow PAMP perception (Fig. 5) [108], [109]. In the event that pathogens possess effectors and deliver them into the host cell to suppress PTI, plants initiate the second layer of the immune response called effector-triggered immunity (ETI), with a majority of the effectors being recognized by the intracellular nucleotide-binding domain, LRR-containing receptors (NLRs) (Fig. 5) [110], [111]. Notably, several DEGs encoding RLK/RLPs (e.g., Zm00001eb239020, 252,380 and 399890) and NLRs (Zm00001eb107150, 226,700 and 226690) were identified in the ER-resistant ESLs (Table S11), suggesting that both the PTI and ETI systems participate in ER resistance in maize. Receptor-like cytoplasmic kinases (RLCKs) are central immune kinases in the PRR complex [112]. In particular, the RLCK gene (Zm00001eb124900) and its orthologs were conserved among maize, rice, and wheat fungal resistance QTLs (Fig. 5, Table S19), further validating the reliability of the mined genic resources obtained here.

A working model for ER resistance in maize

Based on the genic mining information from genetic and multi-omics analyses, we propose a working model for ER resistance in maize (Fig. 5). In plant-pathogen systems, PTI confers a milder form of disease resistance when compared to the robust disease resistance conferred by ETI [113]. Despite distinct activation mechanisms and early signaling steps between PTI and ETI, they converge at various downstream outputs, including MAPK cascades, calcium influx, reactive oxygen species (ROS) burst, transcriptional reprogramming, and phytohormone signaling [114]. Consequently, the DEGs responsive to Fusarium or Aspergillus infection were identified across all signaling sectors (Fig. 5, Tables S8, S11).

MAPK cascades are pivotal signaling hubs in both the PTI and ETI pathways, with the key distinction being that the activation of MAPK cascades exhibits a slower yet more sustained response in ETI than in PTI [115], [116]. A known MAPK (ZmMAPK3) was experimentally confirmed to confer ER resistance (Fig. 5) [65]. The other fundamental components of MAPK cascades, including MAPK (Zm00001eb397190), two MAPK kinases (MAPKK) (Zm00001eb009250 and 377050), and three MAPK kinase kinases (MAPKKK) (Zm00001eb042350, 294,620 and 360950), were identified as DEGs (Fig. 5, Table S8), further validating their significance in ER resistance.

TFs are the main substrates of MAPKs, highlighting the involvement of immune-related MAPKs in transcriptional reprogramming during defense responses [117]. WRKY TFs are one of the main type of substrates in regard to MAPKs [118]. In Arabidopsis, WRKY33 is phosphorylated by MAPKs (MPK3 and MPK6) and plays an essential role in disease resistance [119], [120]. MPK3 and MPK6 can regulate ET biosynthesis through the phosphorylation of ACS2 and ACS6 [121], and WRKY33 directly binds to the promoters of ACS2 and ACS6, indicating that expression of ACS2/6 are dependent on WRKY33 [122]. Several DEGs encoding WRKY TFs (e.g., Zm00001eb149300, 154960, 310260, 337,290 and 368640) located in ER-resistant ESLs were also enriched (Fig. 5, Table S11). Notably, the maize ortholog of WRKY33 (Zm00001eb286490) was among these DEGs, thus opening the door to future investigations into its potential contribution to ER resistance alongside ZmACS2 and ZmACS6 [96]. Respiratory burst oxidase homolog (RBOH) is a plant NADPH oxidase that regulates ROS signaling [123], [124]. In Nicotiana benthamiana, four WRKY TFs can be phosphorylated by MAPK to regulate RBOH during PTI and ETI, indicating that the MAPK-WRKY pathway is required for ETI and PTI ROS bursts by activating RBOH [125]. However, the direct and indirect regulation of MAPK cascades by ROS signals has also been observed [85].

In the nucleus, defense-related genes (DRGs) such as Zm00001eb013990, 167690, and 032,560 are activated by TFs. Additionally, phytohormone-related genes (PRGs), particularly ET-(Zm00001eb074930, 193360, 249,290 and 433310) and JA-(Zm00001eb035010, 336,760 and 406100) related genes, were induced by TFs (Fig. 5). Subsequently, defense-related components accumulate through the activation of related downstream genes (Fig. 5).

At least ten non-race-specific genes that confer broad-spectrum resistance against multiple fungal pathogens have been identified in crops (Table S20). Except for the paralogs of Lr47, Pi9, PigmR, Ptr, Yr36, and TaPsIPK1 which encode typical NLRs or RLCKs, the homologs of Lr34/Yr18, pi21, bsr-d1, and ROD1 were also identified as DEGs (Table S8), indicating that these types of genes also participate in ER resistance in maize. ROD1 encodes a C2 domain Ca²⁺ sensor that promotes ROS scavenging by stimulating catalase activity [126]. One RBOHs, two superoxide dismutases (SODs), and six C2 domain-containing proteins were identified as DEGs (Table S11). Notably, Zm00001eb107970 (C2 domain Ca²⁺ sensor) and Zm00001eb394700 (SOD) were found to be localized in ESLs (Fig. 5, Table S11), suggesting that they play pivotal roles in conferring ER resistance by mediating ROS homeostasis in maize. Interestingly, the Ca²⁺-dependent protein kinase (CPK) OsCPK18, another type of Ca²⁺ sensor, exhibits reciprocal phosphorylation with OsMPK5 in rice, and mutations in the MAPK phosphorylation motif of OsCPK18 and its paralog OsCPK4 simultaneously increase rice yield and immunity simultaneously [127], [128]. In addition to the DEGs that are involved in MAPK cascades, we identified two CPKs (Zm00001eb072810 and 140600) located in ER-resistant ESLs as DEGs (Fig. 5, Tables S8, S11). It would be intriguing to investigate whether the modulation of these Ca²⁺ sensor-MAPK phosphorylation pathways can achieve a trade-off between yield and defense mechanisms in maize.

Glutathione (GSH) is a metabolite that plays an important role in plant responses to biotic stress by removing ROS [129]. Notably, the GSH metabolism pathway was significantly enriched (32 DEGs), with five DEGs locating in ER-resistant ESLs (Fig. 5, Tables S10, S11), indicating that GSH metabolism-related genes may play an important role in the ROS/GSH balance.

The reinforcement of physical barriers, such as the cell wall, represents one of the late defense response events to microbial infection and it serves as an effective passive defense mechanism to restrict the dissemination of infectious agents [130], [131]. Pectin and lignin are the main components of plant cell walls, and the related enzymes (such as pectin methylesterase inhibitor proteins) are essential for antifungal activity and basal disease resistance [132]. Additionally, laccases, along with dirigent proteins and peroxidases, serve as mediators of lignin polymerization [133]. Among over 30 cell wall-related DEGs (Table S9), four participating pectin and lignin synthesis are located in ER-resistant ESLs (Fig. 5, Table S11), indicating that these DEGs play pivotal roles in the primary defense against fusarium or aspergillus infection.

Taken together, the majority of fundamental components in both the PTI and ETI systems have been identified based on our data mining. A significant proportion of these DEGs were located within ER-resistant ESLs in maize. More importantly, many fungal resistance QTLs are located in the syntenic blocks of ER-resistant ESLs in the rice and wheat genomes. These findings provide crucial genic resources for validating gene function, constructing the molecular regulatory network of ER resistance, and enhancing the resistance of crop varieties to diverse fungal infections. However, due to the limitation of current data, it remains challenging to empirically validate the proposed working model (Fig. 5), thus further investigations are needed in the future experimental research.

Genetic linkage or co-localization of genetic loci to regulate both ER resistance and crop yield

The growth–defense trade-off refers to the reciprocal relationship between plant growth and defense [134], [135]. Optimizing this delicate balance has the potential to enhance crop yields under fluctuating environmental conditions. Although the precise molecular mechanism underlying this trade-off remains elusive, accumulating evidence has suggested that limited resource redistribution and positive co-regulation of growth and immunity are the primary driving forces of the growth-defense trade-off [134], [135], [136].

Combined analyses of ER-related genetic and multi-omics data revealed a significant enrichment of DEGs in plant growth-related pathways, particularly those associated with carbon metabolism (Fig. S4, Table S10). It would be intriguing to explore the correlation between defense and yield by utilizing genetic hotspot information on maize yield-related traits [59], [137]. Surprisingly, 71.01 % (120/169) and 80.72 % (67/83) of the QTL and QTN hotspots related to kernel-related traits, respectively, were shown to exhibit overlapping patterns with ER-resistant ESLs in terms of their genomic regions (Table S21). Furthermore, among the 129 kernel trait-regulating genes identified, 21 are located within ER-resistant ESLs (Fig. 6A, Table S22). Regarding ear traits, 49.67 % (76/153) of the QTL hotspots and 30.59 % (26/85) of the QTN hotspots overlap with ER-resistant ELSs (Table S23), and 10 of the 48 known ear-trait genes are localized within ER-resistant ELSs (Fig. 6A, Table S24). Notably, 43 intervals concurrently encompassing hotspots of kernel- and ear-related traits overlapped with ER-resistant ESLs (Tables S21, S23), and 34 and 10 intervals that were detected as QTL/QTN hotspots of kernel- and ear-related traits, respectively, overlapped with ER resistance ESLs (Fig. 6A), implying their potential pivotal roles in regulating both defense and yield in maize.

Fig. 6.

Comparison of the fungi resistance and yield-related genetic loci in crops. (A) Co-locations of the maize ER resistance ESLs with yield related hotspots and known genes. The 34 mutually verified QTL/QTN hotspots by both QTL mapping and GWAS and 21 known genes for kernel traits are labeled in blue, the 10 mutually verified QTL/QTN hotspots by both QTL mapping and GWAS and 10 known genes for ear traits are highlighted in red. Additionally, the regions exhibiting simultaneous co-localization of ER resistance ESLs with hotspots of kernel and ear traits are marked in dark green. (B) Co-locations of the wheat FHB resistance loci with yield related genes. The 38 wheat yield-related genes located within FHB-resistance loci were labeled in red, while the corresponding FHB-resistance loci were labeled in blue.

To further explore whether a similar genetic linkage or co-localization of genetic loci was also present in other crops, 111 and 33 yield-related genes in wheat and rice, respectively, were retrieved (Table S25). We found that 38 wheat yield-related genes were located within the FHB-resistant loci (Fig. 6B, Table S25), indicating that genetic linkage or co-localization of genetic loci controlling both yield and resistance also exists in wheat. However, no such association has yet been observed in rice, possibly due to the limited number of mapped RBD resistance loci.

The identification of yield-resistance loci can provide valuable resources for elucidating the causal genes that regulate the trade-off between crop yield and defense. A major QTL qKW9 was previously reported to affect maize kernel width [138], and was found to co-localize with an FER QTN hotspot (Table S21). The causal gene of qKW9 encodes the DYW motif pentatricopeptide repeat (PPR) protein [139]. It should be noted that PPR proteins have been implicated in the regulation of cell death responses, ROS accumulation, and resistance to fungal and bacterial pathogens in rice [140]. Therefore, it would be intriguing to investigate whether qKW9 confers enhanced ER resistance. The ear trait-associated genes Grx5 and Ead1, which encode a glutaredoxin and an aluminum (Al)-activated malate transporter, respectively [141], [142], are located within the genomic regions overlapping the yield-ER resistance loci (Fig. 6A, Table S23). Both genes participate in ROS homeostasis and regulate ear length in maize [143], [144]. When considering the significance of ROS signaling during PTI and ETI (Fig. 5), it was intriguing to investigate the role of these genes in ER resistance. In a previous study, four significant QTLs that simultaneously regulated both yield and Mediterranean corn borer resistance were identified [143]. Notably, two of these QTLs (bins 5.04 and 8.04–8.05) overlap with the loci associated with yield-ER-resistance (Fig. 6A, Table S23). Remarkably, Grx5 and the functionally known gene Ndl1 related to the yield are located within these genomic regions. Ndl1 encodes an ATP-dependent metalloprotease that regulates ROS accumulation and auxin-related processes [144]. These findings further emphasize the importance of the identified yield-ER resistance loci and their candidate genes, especially those related to ROS signaling, in the regulation of crop yield and also disease resistance.

To the best of our knowledge, this review presents compelling evidence of a genome-wide correlation between loci associated with fungal resistance and yield across diverse crop species. However, it remains unclear whether this significant association arises from the pleiotropic effects of a single gene or the genetic linkage of independent genes, as both genetic models governing crop defense and yield balance have been identified. For example, IDEAL PLANT ARCHITECTURE 1 (IPA1) encodes a TF that plays a crucial role in rice architecture and grain yield [145]. Infection with the fungus Magnaporthe oryzae leads to a shift in its DNA-binding specificity towards the target gene WRKY45, resulting in an enhanced immune response. After 48 hpi, IPA1 was seen to revert to its non-phosphorylated state and resumed growth promotion. Thus, through its pleiotropic effects, IPA1 simultaneously enhances yield and disease resistance [146]. In maize, the auxin-regulated protein ZmAuxRP1 responds rapidly to F. graminearum infection by transiently coordinating root growth with stalk rot and ER resistance [97]. ZmAuxRP1 participates in the biosynthesis of indole-3-acetic acid (IAA) and benzoxazinoid defense compounds and likely acts as a resource regulator to balance growth defense in a timely manner for optimal plant fitness [97].

Another example of genetic linkage is derived from the Pigm locus of rice, which encodes two NLRs that confer broad-spectrum blast disease resistance without negatively impacting yield. PigmR provides fungal resistance, whereas PigmS competitively attenuates PigmR homodimerization to suppress resistance but increase grain yield [147].

Recently, a set of 490 pairs of orthologous genes that underwent convergent selection during the evolution of maize and rice was identified at the genome-wide level. These genes are significantly enriched in pathways related to grain yield, with the majority located within the syntenic blocks between the maize and rice genomes, thus presenting promising targets for future crop improvement [148]. Among these identified genes, 46 (9.39 %) maize genes were located in ER-resistant ESLs (Fig. 7A, Table S26), surpassing what would be expected by random chance when compared to the total pool of 40,621 protein coding genes (1.21 %) in the maize genome [149]. Furthermore, 33, 35, 37, and 38 orthologous genes were present in the syntenic blocks of rice and wheat A, B, and D genomes, respectively (Fig. 7A, Table S26). These genes encode TFs, kinases, and auxin efflux carrier family proteins that participate in carbohydrate, lipid, and amino acid metabolism, signaling, and cellular (such as autophagy) processes (Fig. S6, Table S11). For instance, Zm00001eb075390 encodes autophagy-related protein 13c (ZmATG13c), and one and three orthologous genes represent the syntenic regions of the rice and wheat genomes, respectively (Fig. 7B, Table S26). Autophagy is a highly conserved process that delivers dysfunctional cellular components to vacuoles or lysosomes for recycling during developmental processes and stress responses [150]. Phytohormones, including auxin, GA, BR, JA, and SA, can regulate autophagy at both the translational and post-translational levels [151]. The presence of ATG protein-encoding genes in convergently selected genomic regions in different crops suggests that autophagy may serve as a potential pathway for enhancing fungal resistance and crop yield. In addition, among the 46 maize genes, the expression levels of 34 varied in response to fungal infection (Table S27), highlighting the potential of these genes and their orthologs as targets for modulating the intricate balance between defense and yield in crops.

Fig. 7.

Identification of the 46 maize genes located in ER-resistance ESLs and their orthologous counterparts that were convergently selected during crop evolution. (A) Chromosomal distributions of the 46 maize genes underwent convergent selection and their syntenic relationships with orthologous genes in wheat and rice genome. (B) An example set of syntenic blocks that have potential in regulating crop yield and resistance. The maize ATG13c gene and its orthologs in wheat and rice were visually annotated by red bars. (C) Comparsion of the cross-population composite likelihood ratio (XP-CLR) values between the 46 potential yield and resistance (R/Y) related genes and the remaining 444 yield (Y) related genes.

A severe reduction in nucleotide diversity often occurs in selected regions during crop evolution. This phenomenon has also been observed in regions that have undergone convergent selection [148]. However, as crops must adapt to various hostile habitats, there may be less selective pressure on defense-related genes than on those associated with yield traits [152]. Therefore, the selection rates of the 46 genes and the remaining 444 genes were compared based on cross-population composite likelihood ratio (XP-CLR) levels. The findings revealed that the mean XP-CLR value of the 46 genes was significantly lower than that of the remaining genes (Fig. 7C), supporting our speculation that these 46 genes play important roles in adapting to unfavorable environments.

A proposed regulatory model for trade-off between maize yield and ER resistance

Plants employ photosynthesis to convert luminous energy into chemical energy in the form of carbohydrates, which are subsequently allocated to growth or defense processes contingent on the presence or absence of specific stresses. The activation of plant defense systems places a significant strain on resources such as energy and substances required for plant growth [135]. Therefore, a delicate balance must be maintained between growth and defense to optimize plant fitness and ensure crop yield. Increasing evidence supports the mechanism by which phytohormonal crosstalk plays a central role in regulating the trade-off between plant defense and crop yield [134], [135], [153], [154].

ER defense-related phytohormones included ABA, ET, JA, and SA, but plant growth-related phytohormones, such as auxin, BR, CK, and GA, were also detected as ER-responsive DEGs (Table S8). Based on previous reports and the genic resources identified here, we have proposed the biosynthesis and signaling pathways of these phytohormones that potentially govern resource allocation to maintain a balance between plant defense and grain yield when maize is infected by Fusarium or Aspergillus (Fig. 8). Notably, at least 15 phytohormone-related DEGs involved in auxin, BR, CK, GA, ABA, ET, JA, and SA were located within genetic loci that regulate both ER resistance and maize yield (Fig. 8, Table S28). Specifically, three (Zm00001eb067610, 154930 and 075520) of the 46 genes under convergent selection (Table S26) were auxin-, ABA-, and JA-related genes, indicating their potential involvement in enhancing maize yield and conferring ER resistance through phytohormone-dependent pathways. In addition, convergent selection analysis showed ZmATG13c is a candidate gene that simultaneously regulates maize yield and ER resistance (Fig. 7B, Table S26). These findings indicate that phytohormone-regulated resource recycling pathways may play potential roles in balancing the growth-defense trade-off (Fig. 8). However, owing to the limited number of functionally validated genes, further investigation is required for elucidating the molecular mechanisms underlying these phytohormone pathways in regulating the trade-off between maize yield and ER resistance.

Fig. 8.

A proposed molecular network regulating ER resistance, and the gene resources optimizing maize resistance and yield in breeding programs. Phytohormones regulate resource allocation by modulating the carbon cycling and phenylpropanoid pathways to maintain a dynamic balance between plant defense and grain yield; an ideotype of crops could be generated upon achieving an optimal balance between these two pathways. Candidate genes associated with phytohormones are indicated in parentheses, while candidate genes involved in substrate synthesis are underlined. Solid arrows indicate direct effects, and dashed arrows indicate indirect effects. 4CL, 4-coumarate-CoA ligase; ACO, aconitase; ALDOA, aldolase; ANR, anthocyanidin reductase; ANS, anthocyanin synthase; C3′H, p-coumaroyl shikimate 3′ hydroxylase; C3H, coumarate 3-hydroxylase; C4H, cinnamic acid 4-hydroxylase; CAD, cinnamyl alcohol dehydrogenase; CCoAOMT, caffeoyl CoA 3-O-methyltransferase; CCR, cinnamoyl-CoA reductase; CHI, chalcone isomerase; CHS, chalcone synthase; COMT, caffeate/5-hydroxyferulate3-O-methyltransferase; CSE, caffeoyl shikimate esterase; CSY, citrate synthase; CYP450, Cytochrome P450; DFR, fihydroflavonol 4-reductase; ENO, enolase; F3′5′H, flavonoid 3′5′-hydroxylase; F3′H, flavonoid 3′-hydroxylase; F3H, flavanone 3-hydroxylase; F5H, ferulate 5-hydroxylase; FBPase, fructose-1,6-bisphosphatase; FLS, flavonol synthase; FNS, flavone synthase; FUM, fumarase; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GT, Glycosyltransferase; HK, hexokinase; HCT, Hydroxycinnamoyl-CoA shikimate/quinate hydroxycinnamoyl transferase; IDH, isocitrate dehydrogenase; LAR, leucoanthocyanidin reductase; MDH, malate dehydrogenase; ODC, oxoglutarate dehydrogenase complex; PAL, phenylalanine ammonialyase; PDC, pyruvate dehydrogenase complex; PFK, phosphofructokinase; PGAM, phosphoglycerate mutase; PGI, phosphoglucose isomerase; PGK, phosphoglycerate kinase; PK, pyruvate kinase; PRK, phosphoribulose kinase; RPE, ribulose-5-phosphate 3-epimerase; RPI, ribose 5-phosphate isomerase; RT, rhamnosyl transferase; SCoAL, succinyl-CoA ligase; SDH, succinate dehydrogenase; SPB, sedoheptulose-1,7-bisphosphatase; TKL, transketolase; TPI, triosephosphate isomerase; UFGT, UDPG-flavonoid glucosyltransferase.

Carbohydrates are the main resources for plant growth and crop yield, whereas defensive substrates are mainly synthesized through phenylpropanoid pathways. ER-responsive DEGs were significantly enriched in both the carbon and phenylpropanoid metabolism pathways (Table S12). The phenylpropanoid pathway is responsible for the biosynthesis of a diverse range of phenolics, flavonoids, and lignin that have been implicated in conferring resistance to both biotic and abiotic stresses [155], [156], [157]. Moreover, these processes are highly regulated by phytohormones, including auxin, ET, JA, and GA [155]. We found that DEGs encoding important enzymes in glycolysis (phosphofructokinase, Zm00001eb293530; pyruvate kinase, Zm00001eb066190), tricarboxylic acid (TCA) cycle (isocitrate dehydrogenase, Zm00001eb158070), general phenylpropanoid pathway (phenylalanine ammonialyase, Zm00001eb247660), flavonoid pathway (Glycosyltransferase, Zm00001eb122610) and lignin pathway (CAD cinnamyl alcohol dehydrogenase, Zm00001eb070170), locate within the genetic loci that regulate both ER resistance and maize yield (Fig. 8, Table S28). These findings further corroborate the importance of these pathways in regulating the trade-off between maize yield and ER resistance.

Plants have evolved intricate mechanisms to maintain a delicate balance between growth and defense [147]. Multiple regulatory mechanisms have been reported including phosphorylation variations in the TF IPA1 or the MAPK-CPK regulatory circuit [127], [146], suppression of ETI by PTI priming via the MAPK-WRKY module [158] and epigenetic regulation of paired antagonistic NLRs [147]. Therefore, based on previous reports and the novel information acquired in this review, further research is of great significance to comprehensively elucidate the molecular regulatory mechanisms underlying the trade-off between maize yield and disease resistance and to explore the pivotal regulatory factors used to simultaneously improve crop resistance and yield in future breeding programs.

Conclusions and future perspectives

Environmentally stable QTL/QTN systems play crucial roles in facilitating molecular breeding and candidate gene discovery in crops. However, all three ER types are regulated by a large number of small and unstable QTLs/QTNs [7], [33], [35], which significantly hinders their breeding applications for improving ER resistance in maize. Consequently, severe annual yield reductions due to ER have been often reported worldwide [5], [6], [7]. In this review, 162 ESLs were refined by combining previously reported QTL and QTN information, which would be valuable for elucidating the genetic basis underlying ER resistance and identifying causal genes related to ER resistance for molecular marker development in maize breeding programs. The successful identification of genetic hotspots for multiple quantitative traits validates the efficacy of the strategies employed to mine stable QTLs [47], [59], [159], [160]. Furthermore, although further testing is required, the molecular markers identified in the ESLs and validated by previous studies hold potential for enhancing maize breeding programs.

To date, only 12 maize genes have been functionally validated in regard to ER resistance, leaving the molecular mechanisms underlying ER resistance largely unexplored. Multi-omics analysis revealed a high degree of conservation among these genes in response to Fusarium and Aspergillus infections (Fig. 3B). The derived set of 1391 DEGs, as well as the homologs of known genes that confer broad-spectrum resistance against multiple fungal pathogens, provide valuable genic resources for exploring the molecular mechanisms controlling ER resistance. Furthermore, 297 DEGs located in ESLs (Table S11) provided clues for candidate gene identification and functional validation.

FHB and RBD are the main fungal diseases that pose global threats to wheat and rice production, respectively [14], [16], [18]. An important discovery of this study is that the genetic basis of ER, FHB, and RBD resistance is relatively conserved among maize, wheat, and rice. This conclusion is supported by the fact that homologous genes, such as FHB1 and ZmFER1, have similar functions in fungal infection between wheat and maize [14], [15], [98] and that many QTLs/QTNs for ER, FHB, and RBD resistance are located in the syntenic regions of the maize, wheat, and rice genomes, respectively (Fig. 4C). Therefore, the isolation of ER resistance genes and the dissection of their molecular mechanisms will provide valuable insights for multiple crop species.

Notably, a substantial proportion of ER-resistant genetic loci overlap with yield-related QTLs, QTNs, or genes, and genetic linkage or co-localization of resistance-related genetic loci with yield-related genes has also been observed in other crops (Fig. 6). Although it remains uncertain whether this genetic linkage or co-localization is governed by single pleiotropic genes or multiple independent genes, accumulating evidence indicates that certain hub genes can mediate the growth-defense trade-off [134], [135]. In addition, the 46 genes located in ER-resistant ESLs and their orthologs underwent convergent selection during maize and rice evolution (Table S26), suggesting potential pleiotropic roles of these genes in regulating both resistance and yield.

When considering that ER, FHB, and RBD resistance may share a similar genetic basis and molecular mechanisms among maize, wheat, and rice and that a substantial proportion of genetic loci concurrently regulate these fungal diseases and crop yield, a molecular breeding scheme for optimizing resistance and yield in future crop breeding programs is proposed (Fig. 9). The primary and most efficient approach is the optimal use of known genes and their orthologs (Fig. 9A) as a set of genes mediating plant growth–defense trade-offs have been identified in multiple crops, such as AuxRP1 and ZmFER1 in maize [97], [98], TaPIP2;10, Lr34 and TaPsIPK1 in wheat [161], [162], [163], Pigm, Rod1, IPA1, and OsRLCK176 in rice [126], [146], [147], [164]. For instance, the wheat Lr34 gene, which encodes an ATP-binding cassette (ABC) transporter, has been successfully utilized to confer resistance against multiple fungal pathogens in barley, rice, and maize without detrimental effects on plant growth and crop yield [161], [165], [166], [167].

Fig. 9.

Molecular breeding strategies for optimizing resistance and yield in future crop breeding programs. (A) Utilization of known genes and their orthologs. (B) Exploration of genic resources by combining genetics, multi-omics, and comparative genomics methodologies in crop research. (C) Implementation of marker-assisted selection, genomic selection, as well as gene editing / transgene technologies for crop improvement. (D) Cultivation of elite crop varieties exhibiting enhanced resistance and yield through comprehensive utilization of existing genic resources alongside advanced technologies.

The molecular regulatory mechanisms underlying most crop traits have not yet been thoroughly elucidated and most plant genes have yet to be experimentally validated. Fortunately, a large number of genetic loci and candidate genes for many important agronomic traits have been discovered in crops by combining genetics, multi-omics, and comparative genomics technologies [59], [137], [168], [169]. Here, we have demonstrated that combining these technologies is more efficient for mining genetic loci and genic resources that can potentially balance plant defense and crop yield (Fig. 6, Fig. 7, Fig. 8, Fig. 9B). In addition, it is noteworthy that single-cell RNA sequencing has emerged as one of the most pivotal omics technologies, enabling the elucidation of complex organ or cell type-specific gene responses and cell–cell communications [170], by which we have successfully identified novel expression profiles of maize root tips upon F. verticillioides infection [171]. Moreover, organ/cell-specific regulation of growth-defense trade-off has been revealed in plants [172], highlighting the potential role of single-cell technology in identifying gene resources that balance crop yield and resistance to ER, FHB, and RBD in future studies.

Another consideration is that although both PTI and ETI components were activated upon pathogen challenge (Table S8), no major ER resistance genes with large effect have been identified in maize. Pyramiding sufficient elite alleles of resistance genes or enhancing their expression levels, coupled with the elimination of susceptibility genes through gene editing or other methods, could potentially serve as efficient strategies to improve ER resistance in maize. Notably, targeting susceptibility genes may offer broad-spectrum and more durable protection compared to enhancing resistance genes [173], [174]. Gene editing has been successfully applied for engineering the susceptibility genes to confer resistance to diverse pathogens in several agriculturally important crops [175], [176], [177]. Hence, identification and exploration of potential susceptibility genes associated with ER resistance could significantly contribute to advancing our understanding in this field. Recent studies have shown that transferring the immune receptor RXEG1 from Nicotiana benthamiana confers FHB resistance in wheat [178], Rice blast infections can occur at any stage of rice development, affecting panicles, leaves and stems [179], ectopic expression of wheat gene Lr34 and Momordica charantia gene McCHIT1 confer blast resistance in rice [165], [180]. These findings suggest that using immune receptors from nonhost plants or other sources to enhance ER resistance is a promising approach. Given that FHB1, FHB7 and RXEG1 can confer resistance to Fusarium infection in wheat [15], [16], [178], it is reasonable to test their potential for increasing ER resistance in maize.

MAS has been proven successful in improving agronomic traits controlled by one or a few major effect genes in various crops [181], [182]. However, GS has emerged as a superior approach for selecting agronomic traits influenced by multiple small-effect QTLs [183], [184]. Both methods can also be employed to harness the pleiotropic QTLs governing yield and resistance, regardless of whether the pleiotropism is regulated by a single gene or multiple tightly linked independent genes (Fig. 9C). Because both MAS and GS are time-consuming in breeding programs, important pleiotropic QTLs, particularly candidate genes that have undergone convergent selection during crop evolution, can be quickly functionally validated and utilized through genome editing and transgene technologies (Fig. 9C) [185], [186], [187], [188], [189], [190], [191], [192]. Finally, we envision that through the extensive acquisition of genetic and genic information, coupled with the application of advanced biotechnologies [193], [194], [195], an 'ideotype' for diverse crops exhibiting both high yield and wide spectrum defense will be developed in the future (Fig. 9D).

Compliance with Ethics Requirements

This article does not contain any studies with human or animal subjects.

CRediT authorship contribution statement

Zechao Yin: Data curation, Formal analysis, Visualization, Writing – original draft. Xun Wei: Methodology, Visualization, Writing – review & editing. Yanyong Cao: Data curation, Formal analysis. Zhenying Dong: Conceptualization, Writing – original draft. Yan Long: Funding acquisition, Data curation, Visualization, Writing – original draft. Xiangyuan Wan: Conceptualization, Supervision, Funding acquisition, Project administration, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Biographies

Zechao Yin is currently a PhD candidate in Research Institute of Biology and Agriculture, University of Science and Technology Beijing, China. His research project focuses on exploring genic resources for ear rot resistance in maize.

Xun Wei is a full professor in Research Institute of Biology and Agriculture, University of Science and Technology Beijing, China. Her research focuses on cross-talk between crop genetics & breeding and bioeconomic research in maize.

Yanyong Cao is an associate professor in Institute of Cereal Crops, Henan Academy of Agricultural Sciences, China. His current study mainly concentrates on dissecting molecular mechanisms underlying ear rot resistance in maize.

Zhenying Dong is an associate professor in Research Institute of Biology and Agriculture, University of Science and Technology Beijing, China. His research focuses on identifying genes and elucidating mechanisms related to ear rot resistance in maize.

Yan Long is a full professor in Research Institute of Biology and Agriculture, University of Science and Technology Beijing, China. Her research focuses on grain size and weight gene identification and mechanism elucidation in maize.

Xiangyuan Wan is a full professor and the president of Research Institute of Biology and Agriculture, University of Science and Technology Beijing, China. His group research is involved in study of maize genetics and genomics, and mainly focuses on gene mining and mechanism dissection of important agronomic traits including resistance to ear rot and grain yield. He managed the article design, revised the manuscript, and provided funding support in this review.

Appendix A. Supplementary material

The following are the Supplementary data to this article: