Increasing Fusarium verticillioides resistance in maize by genomicsassisted breeding: Methods, progress, and prospects

Yufng Xu, Zhirui Zhng, Ping Lu, Ruiqi Li, Peipei M, Jinyu Wu,b, To Li,*, Huiyong Zhng,b,*

a College of Life Sciences, Henan Agricultural University, Zhengzhou 450002, Henan, China

b College of Agronomy, Synergetic Innovation Center of Henan Grain Crops and National Key Laboratory of Wheat and Maize Crop Science, Henan Agricultural University,Zhengzhou 450002, Henan, China

Keywords:Maize (Zea mays L.)Fusarium verticillioides Disease resistance Molecular design breeding

ABSTRACT Maize (Zea mays L.) is an indispensable crop worldwide for food, feed, and bioenergy production.Fusarium verticillioides (F.verticillioides) is a widely distributed phytopathogen and incites multiple destructive diseases in maize: seedling blight, stalk rot, ear rot, and seed rot.As a soil-, seed-, and airborne pathogen, F.verticillioides can survive in soil or plant residue and systemically infect maize via roots,contaminated seed,silks,or external wounds,posing a severe threat to maize production and quality.Infection triggers complex immune responses:induction of defense-response genes,changes in reactive oxygen species,plant hormone levels and oxylipins,and alterations in secondary metabolites such as flavonoids, phenylpropanoids, phenolic compounds, and benzoxazinoid defense compounds.Breeding resistant maize cultivars is the preferred approach to reducing F.verticillioides infection and mycotoxin contamination.Reliable phenotyping systems are prerequisites for elucidating the genetic structure and molecular mechanism of maize resistance to F.verticillioides.Although many F.verticillioides resistance genes have been identified by genome-wide association study, linkage analysis, bulkedsegregant analysis, and various omics technologies, few have been functionally validated and applied in molecular breeding.This review summarizes research progress on the infection cycle of F.verticillioides in maize,phenotyping evaluation systems for F.verticillioides resistance,quantitative trait loci and genes associated with F.verticillioides resistance,and molecular mechanisms underlying maize defense against F.verticillioides, and discusses potential avenues for molecular design breeding to improve maize resistance to F.verticillioides.

1.Introduction

Maize(Zea mays L.), also called corn,is cultivated for food, animal feed, bioenergy, and industrial raw materials [1,2].According to the Food and Agriculture Organization of the United Nations(https://www.fao.org/-faostat/zh/#data),the global harvested area of maize in 2021 was 207 Mha,with a total production of 1219 Mt.This output represents 44% of the world’s main grain production,far surpassing that of rice and wheat.As population expands,crop production would have to increase by 70% to meet the demand of an expected population of 9 billion by 2050 [3].However, maize production and quality is often limited by fungal diseases, such as late wilt caused by Magnaporthiopsis maydis [4], southern corn rust caused by Puccinia polysora [5,6], banded leaf and sheath blight caused by Rhizoctonia solani [7], head smut caused by Sporisorium reilianum[8],southern leaf blight caused by Cochliobolus heterostrophus[9],and gray leaf spot caused by Cercospora zeaemaydis [9].

Fusarium verticillioides (F.verticillioides), previously known as F.moniliforme (teleomorph Gibberella moniliformis, also reported as mating population A of the G.fujikuroi complex)[10],is a common fungal pathogen that infects many plant species including maize,wheat, rice, sorghum, millet, and North American native grasses[11].F.verticillioides typically grows as an endophyte in plant tissues without causing visible symptoms, but under conducive weather conditions or stress from insect or fungal attack,can cause severe symptoms of seedling blight, stalk rot, cob rot, ear rot, and seed rot[12].F.verticillioides contamination was found in 51 of 135 cereal samples collected in southern India [13].In rice, 11.6% of 225 fungal strains collected from plants with seedling blight were identified as F.verticillioides [14].In wheat, Fusarium-caused head blight(FHB)has become a disastrous disease worldwide,especially in China, and has expanded from the middle and lower Yangtze valleys to the entire Yellow and Huai River valleys in recent decades[15].In maize,F.verticillioides is one of the most common fungal pathogens [16–20].Fusarium stalk rot caused yield reductions of up to 30%–50% in maize [16,21].F.verticillioides-caused Fusarium ear rot(FER)is widespread throughout the global maize cultivation areas [22].In China, the incidence of FER typically ranges from 5% to 10% in normal years, but in heavy outbreak years, the incidence can reach 30%–40% [23].In some susceptible cultivars,the incidence can be over 50%, resulting in a yield loss of at least 30% [23].In Europe, FER is estimated to cause yield losses of 10%–30% [24].The issue of FER is particularly severe in maize–wheat rotation systems because F.verticillioides surviving in maize residues can lead to FHB in wheat [25].

Besides reducing crop yield and quality, F.verticillioides produces fumonisins, a group of mycotoxins harmful to human and animal health and classified as probable carcinogens [13,26].High concentrations of fumonisin are found in F.verticillioides-infected plants or in stored seeds, especially when wet and warm weather persists before harvest [27,28].Fumonisin-containing food or forage has been linked to neural-tube birth defects in humans,as well as neurotoxicity, hepatotoxicity, and nephrotoxicity in animals[13,29–31].

Despite crop rotation and irrigation measures, biocontrol agents, and application of chemical fungicides have been used to reduce F.verticillioides-caused maize disease,but are neither effective nor environment-friendly [32–35].The preferred method for controlling F.verticillioides infections and reducing mycotoxin levels is to breed resistant maize cultivars [18,27].To achieve this goal, the development of a reliable phenotypic evaluation system,dissection of physiological and biochemical characteristics of maize responsive to F.verticillioides, and identification of resistant maize germplasm resources and resistant quantitative trait loci(QTL) and genes are essential steps for improving breeding of maize resistance to F.verticillioides.

Although QTL mapping, genome-wide association study(GWAS)and omics research have identified a series of QTL or genes associated with F.verticillioides resistance[18,36–40],the molecular mechanism underlying maize resistance to F.verticillioides remains unclear.Here we review the F.verticillioides infection cycle in maize,phenotypic evaluation for F.verticillioides resistance,QTL and genes associated with F.verticillioides resistance, and progress in identifying molecular mechanisms underlying maize defense against F.verticillioides,and discuss possible approaches for molecular design breeding to improve maize resistance to F.verticillioides.

2.F.verticillioides infection cycle in maiz e

Fig.1.Fusarium verticillioides infection cycle in maize.As a soil-,seed-,and airborne pathogen,F.verticillioides can systemically infect maize via root,contaminated seed,silks,external wounds,stomata,or trichomes,causing asymptomatic endogenous infection or seedling blight,stalk rot,ear rot,and seed rot.After harvest,the fungus can survive on plant residues,aggressively proliferate in roots and mesocotyl,and then invade maize stems through the vasculature,providing inoculum for a subsequent infection cycle.

A comprehensive understanding of the F.verticillioides infection cycle is crucial for developing effective control strategies.F.verticillioides is a soilborne fungal pathogen that can survive in maize residue for up to 630 days,both on the soil surface and in buried stalks at depths of 15 or 30 cm[41],and surviving F.verticillioides in maize residues can readily colonize soil(Fig.1)[42].Germination of seeds in F.verticillioides-infested soils, or of seeds with F.verticillioides present inside them or on their surface,may lead to poor germination rates,uneven height and weight of emerging plants,and potentially aggressive root rot, seedling blight, or seedling death,reducing plant population density and stand uniformity (Fig.1)[43–46].In some cases,contaminated seeds may develop no symptoms after germination but already carry endophytic infection[46].Under favorable conditions, endophytic F.verticillioides aggressively proliferates in roots and mesocotyls and then invades maize stems through the vasculature, causing stalk rot (Fig.1) [46–48].Stem wounds caused by mechanical damage or insect feeding become infection sites of external F.verticillioides and may also lead to stalk rot [46,49].At the silking stage, airborne F.verticillioides spores germinate on maize silks, after which fungal mycelia grow down the silks and cause kernel infection with a‘‘starburst”appearance[18,46].Kernel infection was greatly facilitated by mechanical wounding created by insect feeding, birds, or hail damage (Fig.1)[50–52].Kernel infection can also arise from systemic infection,in which seedborne or soilborne F.verticillioides propagates inside the young plant,spreading from roots to stalk and ultimately infecting the maize cob and kernels [10,53].Stomata and trichomes of maize leaves provide an entry point for infection by airborne F.verticillioides, potentially leading to cob and kernel infection [54,55].Kernel infection can develop into ear rot,seed rot,or asymptomatic endophytic infection[10].During seed storage,seedborne external F.verticillioides or endophytic infection may develop into visible rot symptoms in favorable environments(Fig.1),reducing seed storage time,seed vigor,and germination rate[56],and even causing a new cycle of systemic infection[10,46].

3.Phenotypic evaluation system for F.verticillioides resistance in maize

The establishment of a precise and convenient phenotypic evaluation method for F.verticillioides resistance,capable of accommodating large populations, is a prerequisite for conducting genetics research on F.verticillioides-caused diseases[27].However,achieving this goal is challenging,owing to numerous factors influencing disease resistance scoring: inoculation time, inoculation method,and choice of quantitative indicator of disease severity [56].Although some maize cultivars exhibit obvious symptoms of F.verticillioides-caused rot without artificial inoculation, the unstable disease indices resulting from heterogeneous pathogen exposure across genotypes under natural condition makes it difficult to distinguish resistance variation among cultivars [37,57].Artificial inoculation is necessary to ensure equal distribution of the pathogen in the field, thus promoting F.verticillioides infection and enabling consistent phenotypic screening [37,56,58].Owing to the large genotype–environment interactions of disease resistance,artificial infections should be conducted across multiple environments and years [36,59,60].The inoculation amount and method must produce infection levels sufficient to distinguish resistance differences among genotypes,but not so severe that the differences are obscured [52].

3.1.Seedling blight

Inoculation for seedling blight can be completed in controllable laboratory conditions(Table 1).There are two common inoculation methods for F.verticillioides-caused seedling blight: seed infection and seedling infection (Table 1).In the seed infection method, a spore suspension is applied to mature kernels on the embryo side,and the effect of F.verticillioides on seedling length,weight,and rot is evaluated using the rolled towel assay phenotyping method[20,44,45,61].However, the inoculation amount varies widely(e.g., from 100 μL 1 × 106to 100 μL 3.5 × 106spore suspension per milliliter) depending on the maize population used[20,44,45].Sterilized seeds can also be directly incubated in spore suspension overnight and then planted in sterile soil to evaluate the effect of F.verticillioides infection on seedling growth [62,63].However, seedling growth is often influenced by variation in ger-mination rate and germination vigor among genotypes,which may interfere with the assessment of F.verticillioides resistance(Table 1).In the seedling infection method, mesocotyls of twoweek-old seedlings are scratched with a syringe needle and a spore suspension is applied directly to the wound site [64–66].Subsequently, the relative F.verticillioides content is quantified, and the decay of mesocotyls is photographed to evaluate the resistance of genotypes to F.verticillioides-caused seedling blight [64–66].

Table 1 Comparison of inoculation methods for evaluating F.verticillioides resistance in maize.

3.2.Stalk rot

Two inoculation methods have been employed for evaluating F.verticillioides-caused stalk rot: wound channel and soil burial(Table 1).In the wound channel method, the stalks of adult plants are punctured with a needle,and the wound sites are treated with a cotton swab soaked in spore suspension [64,67] or a toothpick covered with fungal growth[68,69],or spore suspension is injected directly[70–72].The wound sites are sealed with Parafilm for several days,after which the lesion areas of longitudinally split stalks are measured[64,67–71].This method can overcome plant’s physical barriers and promote disease progression (Table 1).For soil burial, maize stalks are inoculated by burying F.verticillioidesinfected kernels in a hole 5–10 cm from each plant followed by irrigation to increase soil moisture for promoting F.verticillioides growth and infection [73,74].The stalks are split longitudinally to assess the level of stalk damage after a period of growth[73,74].This method better simulates the infection process of soilborne F.verticillioides (Table 1).

3.3.Seed rot

Seed rot can be caused by either internally or externally occurring F.verticillioides [56].In one study [17], healthy seeds were surface-sterilized and incubated in spore suspension in vitro for 12 h, followed by a 7-day incubation period to evaluate the effect of externally occurring F.verticillioides on seed rot.This method can be completed under controllable laboratory conditions, but it is hard to distinguish internal seedborne from external F.verticillioides infection(Table 1).In another study[56],kernels were inoculated in vivo on day 15 after pollination, after which asymptomatic seeds adjacent to the inoculation site were sampled,surface-sterilized, and incubated for five days to evaluate seed rot caused by internally occurring F.verticillioides.This method promotes the infection process of internally seedborne F.verticillioides.However,high variation in inoculation time and field environment can also lead to varying disease severity (Table 1).Although the same maize lines were used, the investigation time varied from five [56] to seven [17] days after incubation.The difference in times for seed rot development may have been due to the direct access of internal F.verticillioides to the nutrients in the embryo and endosperm, whereas external F.verticillioides must penetrate the seed coat to use these nutrients.Some researchers[64,67]have inoculated seeds by directly scratching pericarps and applying spore suspension to the wound site.

3.4.FER

Resistance to FER includes both cob and kernel resistance [19].In the dip cob-stab method [19], a 2-cm-deep, 2-mm-diameter hole was drilled in the middle of the cob and the spore suspension was applied to the hole.The length of infected cob was measured to evaluate cob rot resistance.But most studies do not separate cob resistance from ear resistance.In general, there are two common inoculation methods for evaluating FER: silk channel and wound channel (Table 1).Silk channel inoculation involves spraying the spore suspension onto the maize silks using an atomizer [75] or injecting it into the silk channel near the cob tip with a syringe[59,76–80].This method better simulates the process of plant infection in the natural environment(Table 1).In contrast, wound channel inoculation involves damaging the ears with a needle and injecting the spore suspension directly into the wound site or placing a F.verticillioides-colonized toothpick/sponge against it[36,37,39,60,80–90].Wound channel inoculation mimics the plant infection process that may occur due to insect feeding, bird pecking, or hail damage, and this method can overcome many of the plant’s morphological barriers (Table 1) [18].A comparison [75]of the silk channel and wound channel inoculation techniques showed that only the injection of inoculum through the husk leaves significantly increased the severity of FER compared to the control,possibly explaining why most studies use the wound channel inoculation method.Inoculation for FER is generally performed during the grain filling period,which is typically 15 days after pollination [82].However, high variation in flowering time among genotypes makes it time-consuming to identify a suitable inoculation stage [27], and varying environmental conditions during the inoculation period can also lead to varying disease severity(Table 1).

4.Genomic strategies for identification of QTL and genes associated with F.verticillioides resistance

To identify the molecular mechanisms underlying maize resistance to F.verticillioides and promote marker-assisted breeding of resistant maize cultivars,it is desirable to dissect the genetic architecture of F.verticillioides resistance and identify QTL and genes affecting it.Numerous studies [27,52,60,91] have shown that F.verticillioides resistance is a complex quantitative trait controlled by multiple genes with high heritability.To identify QTL and genes for resistance against F.verticillioides, many genomic and bioinformatics strategies have been developed,including linkage mapping,GWAS, and various omics technologies.

4.1.Linkage mapping

Linkage mapping for F.verticillioides resistance has usually relied on recombinant inbred lines (RILs) derived from biparental crosses (Table 2).In early studies, linkage maps were constructed using a limited number of DNA markers such as restriction fragment length polymorphisms (RFLP) and simple sequence repeat(SSR) markers (Table 2).Owing to the small number of markers and the small population size,the resolution was often low,resulting in QTL confidence intervals spanning from several to dozens of centiMorgans that contained hundreds of genes [18,76,84,92].In one study [76], 143 RILs (NCB, NC300 × B104) and 213 BC1F1:2families (GEFR, GE440 × FR1064) were genotyped with respectively 113 and 105 SSR markers, resulting in average distances of 19.4 and 20.7 cM between markers for the NCB and GEFR maps.Although the GEFR and NCB populations identified seven and five QTL, explaining respectively 47% and 31% of phenotypic variation for FER resistance, these QTL regions spanned large map intervals.Another study[84]using 187 RIL lines(87–1×Zong 3)revealed six QTL for FER based on a genetic map containing 246 SSR markers with average genetic intervals of 9.1 cM.

With the emergence of next-generation sequencing technologies, single-nucleotide polymorphism (SNP) markers have displayed their potential for uncovering genes and QTL associated with F.verticillioides resistance (Table 2).Ma et al.[56] identified six QTL for seedborne F.verticillioides seed rot using 7458 polymorphic SNPs for linkage map construction,and the major QTL,qISFR4-1, located on chromosome 4 in an interval of 495 kb, explained 16.6% of total phenotypic variance.Feng [89] used high-quality SNPs to construct linkage maps for three RIL populations developed by crossing two different teosintes with the maize inbred lines B73 and Zheng 58, and four QTL were identified in two RIL populations.

Besides RIL populations, F2:3populations have been used for linkage mapping (Table 2).Chen et al.[93] identified three QTL for FER using 210 F2:3families (BT-1 × Xi502), among which the QTL located on chromosome 4 (bin 4.05/06) accounted for 18% of phenotypic variation.Based on genotyping by sequencing (GBS),Maschietto et al.[37]constructed a genetic map for 188 F2:3plants(CO441× CO354),identifying 15 QTL for FER,eight of which were also associated with fumonisin contamination.Wen et al.[60]identified 20 FER-resistance QTL using three F2populations [60].qRfer1, qRfer3, and qRfer4 together accounted for 62.9%–82.3% of phenotypic variation.

Although many QTL for F.verticillioides resistance have been identified in biparental populations generated by crossing two homozygous lines,this approach exposes limited genetic variation,as it is restricted to the differences between the two parents [18].To overcome this limitation,multi-parent population designs have been developed in crops to expand allelic diversity and increase the frequency of recombination in RILs [20].Examples a highly diverse nested association mapping (NAM) population developed from crossing 25 diverse inbred maize lines to the B73 reference line[94],and a multi-parent advanced generation intercrossing(MAGIC)population developed from eight founder lines [20,95].Septiani et al.[20]used a MAGIC population consisting of 401 RILs and genotyped them using 56,110 SNPs to conduct high-definition QTL mapping for F.verticillioides seedling rot, identifying three QTL for Fusarium seedling rot.Butrón et al.[85]constructed a MAGIC population using the eight founders A509,EP17,EP43,EP53,EP125,F473,EP86,and PB130,and used 339 RILs and 58,556 SNPs to perform QTL mapping for FER.They identified 13 putative minor QTL[85].

4.2.GWAS

With the development of whole-genome sequencing technology, the use of high-density SNP datasets in combination with GWAS has emerged as a powerful alternative to biparental population-mediated QTL mapping(Table 2).Association mapping can dissect F.verticillioides resistance at the population level by exploiting the high levels of natural diversity and historical recombination events [80].Stagnati et al.[45] performed a GWAS for F.verticillioides seedling blight using 230 lines, identifying 42 SNPs associated with F.verticillioides resistance at the seedling stage and 18 genes containing or adjacent to the identified SNPs co-localized in QTL regions previously reported to be associated with F.verticillioides resistance.Mu et al.[19] used 289,659 SNPs and phenotypic data from 258 lines from Tangsipingtou (TSPT), Reid,and the International Maize and Wheat Improvement Center(CIMMYT), identifying 28 genes associated with 48 SNPs associated with Fusarium cob rot resistance, two of which were located near a major QTL qRcF.verticillioides2 detected by linkage mapping.Zila et al.[78] conducted GWAS for FER in a panel of 1687 diverse inbred lines from the USDA maize gene bank and identified seven SNPs in six genes associated with FER resistance.Coan et al.[80]conducted GWAS using phenotypic data from 183 tropical inbred lines and 267,525 SNPs generated by genotyping by sequencing,identifying 14 SNPs associated with FER.

Table 2 Identified QTL and SNPs associated with F.verticillioides-caused diseases and their experimental details.

4.3.Combination of GWAS and other methods

The combination of GWAS with other techniques such as linkage mapping, bulked-segregant analysis (BSA), and transcriptome analysis has proven to be an effective method for identifying F.verticillioides resistance genes (Table 2).Guo et al.[87] performed GWAS and BSA using 509 diverse maize lines and 37,801 SNP markers,and identified 23 SNPs and various candidate genes associated with FER resistance, of which bin 10.03 was detected by both GWAS and BSA.Combining linkage mapping and GWAS, Ju et al.[17] identified eight QTL and 43 genes associated with 57 SNPs associated with Fusarium seed rot resistance.The function of GRMZM2G0081223, detected by both linkage mapping and GWAS,was further verified by the significantly improved F.verticillioides resistance of the near-isogenic line (NIL) containing the resistant allele of GRMZM2G0081223.Wu et al.[83] identified 18 SNPs associated with FER using GWAS,with five SNPs sharing common intervals with QTL detected by linkage mapping.Chen et al.[36]performed GWAS using 818 tropical maize lines and identified 45 SNPs associated with FER resistance.Linkage mapping in four bi-parental populations was then used to validate GWAS results and showed eight common loci in chromosome bins 2.04, 3.06,4.04, 4.08, 5.03, 5.04, 9.01, and 10.03.Yao et al.[39] combined GWAS and transcriptome analysis and identified 69 genes associated with FER resistance.More than one fourth of the genes detected by the GWAS were differentially expressed between the most resistant and most susceptible lines.

4.4.Combination of multiomics technologies

Although GWAS and linkage mapping can identify genes associated with F.verticillioides resistance,functional verification of these genes still faces great challenges,as F.verticillioides resistances are complex quantitative traits usually controlled by polygenes with minor effects [56].In addition, the locations of identified QTL or genes often conflict among studies, likely owing to environmental influence [18].For this reason, GWAS and linkage analysis require the integration of phenotypic data collected from multiple environments over many years,as well as the expansion of populations and numbers of markers,to increase the reliability of the identified QTL [18].However, this process is time-consuming, laborintensive, and expensive [27].To address these challenges,a combination of multiomics technologies and bioinformatics analyses should be adopted to systematically dissect the mechanisms of defense by genes against F.verticillioides [27,34,96].Zhou et al.[97] used deep sequencing to compare small RNA libraries from F.verticillioides-infected maize kernels of susceptible and resistant lines and identified 251 differentially expressed miRNAs, among which Zma-miR408b directly participated in FER resistance,as verified by transgene evidence.By transcriptomic analysis of F.verticillioides-resistant BT-1 kernels before and after F.verticillioides inoculation, Wang et al.[40] identified 1,480 genes responsive specifically to F.verticillioides infection, among which GRMZM2G139874,GRMZM2G028677,and GRMZM2G075333,participating in phenylalanine metabolism, were increased in F.verticillioides-inoculated samples.By transcriptomic analysis of a F.verticillioides-susceptible African maize line CML144 before and after F.verticillioides inoculation, Lambarey et al.[98] found that the kauralexin biosynthesis protein such as TPS1 and cytochrome P450 were involved in CML144 resistance to F.verticillioides infection.Campos-Bermudez et al.[99] observed transcriptional and metabolic changes after F.verticillioides infection in a resistant(L4637) and a susceptible (L4674) maize line, and found that the levels of betaglucosidase (MZ00044454), beta-1,3-glucanase(MZ00027418) and pectinesterase (MZ00030453) increased in the inoculated grain of the susceptible line L4637,along with a rise in mannitol and sorbitol polyols after inoculation in L4674 but not in L4637.However, successful integration of genomics, transcriptomics,proteomics,and metabolomics for maize response to F.verticillioides has been infrequently reported.

5.Molecular mechanisms of maize response to F.verticillioides infection

5.1.Pathogen-associated molecular pattern-triggered immunity (PTI),effector-triggered immunity (ETI), and F.verticillioides resistance

Throughout the coevolution of plants with natural enemies,plants have developed multiple levels of defense mechanisms against microbial pathogens [100–102].The first line of defense is initiated by pathogen-associated molecular patterns (PAMPs),which are conserved pathogen molecules such as bacterial flagellin, elongation factor Tu (EF-Tu), and fungal chitin [100–102].The perception of PAMPs by membrane-associated patternrecognition receptors(PRRs)leads to PTI,which involves the influx of calcium across the plasma membrane, activation of calciumdependent kinases (CPKs), a burst of reactive oxygen species(ROS), and mitogen-activated protein kinase (MAPK) cascades[102].The second line of plant immunity, referred to as ETI, is dependent on the specific recognition of pathogen effectors by intracellular nucleotide-binding leucine-rich repeat receptors(NLR)and leads to strong immune responses such as the induction of hypersensitive response (HR) [100–102].Lanubile et al.[34,96,103] reported a wide range of genes encoding CPKs, MAPK,G proteins, and NLR proteins that showed altered expression after F.verticillioides infection (Fig.2), some of which were highly induced in the resistant genotype, indicating their involvement in maize resistance to F.verticillioides.Many pathogenesis-related(PR) genes, such as PR1, β-1,3-glucanase, chitinase (PRm3), PR4,PR5, PRm6, and thaumatin-like protein, were up-regulated after F.verticillioides challenge.However, the limited research leaves unclear whether F.verticillioides resistance in maize involves typical PTI and ETI immunity.

5.2.ROS and F.verticillioides resistance

Elevated levels of ROS produced by the plasma membranelocated NADPH oxidase, respiratory burst oxidase homologue(RBOH),serves as an early signal in response to pathogen infections(Fig.2)[104–106].PAMP-induced ROS accumulation was produced by respiratory burst oxidase homologue D (RBOHD) [107], which was activated by surface-localized PRRs upon recognition of several PAMPs (flg22, elf18, and chitin), prior to ROS production[105,108].A wide range of genes encoding calcium-activated protein kinases, such as CPKs and CBL-interacting protein kinases(CIPKs), were induced after F.verticillioides infection[34,96,103,109], which might further activate RBOH protein to induce early ROS production [106,110].Impairment of PAMPinduced ROS burst was observed in the quadruple mutant cpk4/5/6/11 of Arabidopsis,but absence of the Ca2+channels CNGC2 and GLR3.3 also impaired PAMP-induced ROS production[104,111].For biotrophic pathogens, which infect living cells,increased intracellular ROS levels in infected tissues can trigger an immune cascade and promote host immunity[112–115].However,in a necrotrophic pathogen,which prefers dead cells,overproduction of ROS during pathogen infection can cause cell death and tissue necrosis, which favor F.verticillioides invasion by offering a growth substrate,thus increasing host susceptibility[116].F.verticillioides-infected tissues of a susceptible maize line showed higher H2O2levels and more cell death than a resistant line[116,117].The rapidly produced ROS triggers cellular antioxidant systems, and the activities of many antioxidant enzymes, such as superoxide dismutase (SOD), catalase (CAT), peroxidase (POX), glutathione peroxidase (GPX), ascorbate peroxidase (APX), glutaredoxin(GRX), and glutathione-S-transferase (GST), are impaired upon F.verticillioides infection [34].Activities of CAT, APX and POX were increased in both resistant and susceptible lines after F.verticillioides inoculation [70].But APX and SOD activities were higher in resistant than in susceptible seedlings before F.verticillioides infection and remained unchanged five days after inoculation,whereas all enzymes assayed were activated after F.verticillioides inoculation only in susceptible seedlings [117], indicating a basal constitutive defense mechanism provided by the resistant genotype.In agreement with this finding,defense-response genes were transcribed at high levels before F.verticillioides invasion and provided basic defense against the fungus in the kernels of a resistant line, whereas these genes were induced specifically by F.verticillioides infection from a basal level in susceptible kernels [96,103].

Fig.2.Molecular mechanisms of maize response to F.verticillioides infection.

5.3.Plant hormones and F.verticillioides resistance

Plant hormones function in plant systemic immunity against pathogens [34,40].The systemic immune response triggered by local plant–microbe interactions has been classified as systemic acquired resistance (SAR) or induced systemic resistance (ISR)based on the site of induction and the lifestyle characteristics of the inducing microorganism [118].SAR is induced by pathogenic microorganisms and relies primarily on the salicylic acid (SA) signaling pathway, targeting hemibiotrophic and biotrophic pathogens [119].SA-responsive genes are not significantly induced in either resistant or susceptible kernels after F.verticillioides infection [34].In contrast, ISR relies mainly on the jasmonic acid (JA)and ethylene(ETH)signaling pathways[120].In addition to hemibiotrophic and biotrophic pathogens,ISR is effective against necrotrophic pathogens [121].The JA-deficient zmopr7zmopr8 double mutant displays complete lack of immunity to F.verticillioides[64].Maize lipoxygenase (LOX) genes were necessary for JA biosynthesis during maize defense against F.verticillioides infection[122], and conversely, an increase in JA content induced the expression of ZmLOX [64].The reduced resistance of zmlox12 to F.verticillioides may be due to lower levels of the JA precursor 12-oxo phytodienoic acid and JA-isoleucine,as well as reduced expression of jasmonate-biosynthetic genes[64].Although ISR is defined[123] as being triggered by beneficial bacteria in the plant rhizosphere, JA biosynthesis is increased after F.verticillioides infection[64,122].Many typical JA-responsive defense components, such as JA-induced protein, chitinases, LOX and PR10, were strongly activated at 72 h post-inoculation in a resistant line [34].In addition to JA, numerous genes involved in ETH biosynthesis and signaling are differentially expressed between F.verticillioidesresistant and susceptible maize lines: ACC oxidase genes, Sadenosylmethionine synthase (SAMS), ETH-responsive protein(ERF) encoded genes, and ethylene-insensitive 3 putative genes[34,39,124,125].

Other hormones, such as auxin (AUX), abscisic acid (ABA), and gibberellin (GA), participate in maize defense against F.verticillioides[39,40,98,103].Rapid but transient reduction in the expression of ZmAuxRP1, a plastid stroma-localized auxin-regulated protein, suppressed the biosynthesis of indole-3-acetic acid (IAA)and promoted the formation of benzoxazinoid defense compounds,thus resulting in arrested root growth but increased resistance to Gibberella stalk rot and FER [73].A novel maize microRNA negatively regulated resistance to F.verticillioides by targeting ZmGA2ox4, which encodes gibberellin 2-oxidase 4, a regulator of the deactivation of bioactive GAs [116,126].Overexpression of ZmGA2ox4 increased resistance to F.verticillioides-caused seedling blight and seed rot in Arabidopsis[116].Several transcriptional factors including WRKY, MYB, and NAC, which can be activated by hormone signaling,are also responsive to F.verticillioides infection[34,103,124].A 1.5-fold induction on average was observed [34]for all NAC genes and MYBs following F.verticillioides inoculation in CO441 kernels.

5.4.Oxylipins and F.verticillioides resistance

Oxylipins, a group of oxidized lipid molecules, are emerging as signals in disease development and production of spores and mycotoxins [127–129].The synthesis of oxylipins is dependent on LOX, which are commonly classified into two major subfamilies: 9-LOX and 13-LOX, based on the specific carbon atom in the fatty acid chain that is oxygenated [130,131].Maize has 13 LOX genes [132,133], several of which function in the maize-F.verticillioides interaction [134,135].Loss of ZmLOX4 leads to increased susceptibility to F.verticillioides and elevated mycotoxin content in diseased kernels,as well as altered LOX enzymatic activity upon F.verticillioides infection[61,134].ZmLOX12 is strongly induced by F.verticillioides infection and its absence results in reduced accumulation of many kinds of oxylipins and increased susceptibility to F.verticillioides colonization in mesocotyls, stalks, and kernels,as well as a greater amount of fumonisin in infected kernels [64].ZmLOX5 acts as a positive regulator in maize defense against F.verticillioides [122].Conversely, a zmlox3 mutants exhibited reduced levels of several 9-LOX-derived hydroperoxides, resulting in increased resistance against F.verticillioides stalk rot as well as reduced spore production and fumonisin B1 production in F.verticillioides-infected kernels [67,122].The augmented resistance may be linked to a more robust and timely activation of ZmLOX4,ZmLOX5,and ZmLOX12,which interfere with JA biosynthesis during maize defense against F.verticillioides infection [67,122].

5.5.Secondary metabolites and F.verticillioides resistance

The defense response of maize to F.verticillioides is linked to the accumulation of various secondary metabolites, including flavonoids,phenylpropanoids,phenolic compounds,and benzoxazinoid defense compounds [18,34,40].The content of 3-deoxyanthocyanidin flavonoids is induced in the silks and kernels of resistant maize lines upon F.verticillioides inoculation,indicating their critical role in F.verticillioides resistance [136].Correspondingly, many genes involved in flavonoid biosynthesis showed markedly changed expression patterns after F.verticillioides infection[34,40,125],some of which were responsive to F.verticillioides inoculation specifically in a resistant line [34,40,137].The flavonoid biosynthesis gene ZmF3H was upregulated by F.verticillioides,and its promoters contain binding sites for the transcription factors ZmDOF and ZmHSF, which were also induced by F.verticillioides(Fig.2) [125].Pericarp phenylpropanoids served as resistance factors against F.verticillioides, as a resistant maize line with lower levels of ear rot and grain fumonisin concentration after F.verticillioides inoculation exhibited higher levels of phenylpropanoids[138].Induction of some genes involved in phenylpropanoid biosynthesis was more pronounced in a resistant genotype[34,39,40].Zm00001d029218 (ZmNCS), which encodes (S)-norcoclaurine synthase associated with the biosynthesis of benzoxazinoid alkaloids, was induced after F.verticillioides infection[125], highlighting the role of alkaloids in maize defense against F.verticillioides.The content of DIMBOA, a benzoxazinoid defense compound, was higher in seedling roots of the resistant line Y331-R than that of the susceptible line Y331-S 24 h after inoculation, accompanied by differential expression of a series of genes involved in DIMBOA synthesis pathway [73].

The perception of F.verticillioides-associated molecular patterns by membrane-associated pattern recognition receptors (PRRs), as well as the recognition of F.verticillioides-related effectors by intracellular nucleotide-binding leucine-rich repeat receptors (NLR),result in two distinct immune responses: PTI and ETI.These responses involve several cellular processes, including the release of reactive oxygen species (ROS), the entry of calcium ions (Ca2+)through the plasma membrane, the activation of calciumdependent kinases(CPKs),CBL-interacting protein kinases(CIPKs),G proteins,and mitogen-activated protein kinase(MAPK)cascades.Consequently, many defense signal pathways are activated, and this process involves the induction of F.verticillioides defenseresponse genes such as ZmLOXs, ZmGA2ox4, and ZmDOF, causing changes in the accumulation of antioxidant enzymes,defense hormones, oxylipins, and various secondary metabolites.

6.Approaches for improving F.verticillioides resistance in maize

6.1.Germplasm screening for F.verticillioides resistance

Conventional breeding holds great promise for mitigating damage by F.verticillioides to maize yield and quality [56].Its success relies on accurate and efficient phenotypic evaluation of F.verticillioides resistance, identification and screening of resistant germplasm, and successful transfer of the resistance trait to selected recipient lines [139].In recent years, numerous elite maize inbred lines have been identified for their high resistance to F.verticillioides.Ma et al.[56]identified 17 inbred lines with high resistance to seedborne F.verticillioides: CNW133, CML172, CNW035,CNW058, CML141, CNW054, CML171, CML300, CNW134,CNW120, CML140, CNW051, CML40, CNW028, CNW031,CML480, and CNW026, and these lines hold potential as donors for increasing maize resistance to seed rot.Four inbred lines,02C14609,02C14643,02C14654,and 02C14678,have been classified as highly resistant or resistant to stalk rot in multiple locations and years and can be used in the development of F.verticillioides stalk rot-resistant maize lines [69].Maize inbred lines displaying high resistance to FER include CO441 [37], GE440 [76,140],NC300 [76,140], Cheng351 [60], Dan598 [60], JiV203 [60],LP4637 [79], 87-1 [84], CIMBL47 [39], Shen137 [87], BT-1 [93],CML492 [36], CML495 [36], CML449 [36], and Qi319 [59], which can serve as parental lines in conventional breeding to improve FER resistance.

6.2.Marker-assisted breeding

Conventional breeding for maize resistance to F.verticillioides is often characterized by long cycles,heavy workload,and low selection efficiency,which restrict the quality and speed of breeding.In contrast, molecular design breeding, which usually includes marker-assisted breeding, transgenic approaches, and genome editing technologies,is considered an efficient approach for generating maize cultivars resistant to F.verticillioides [27,52,87,141–143].For marker-assisted breeding, an identified F.verticillioidesresistance QTL can be introduced into recipient lines using DNA markers linked to the QTL,even if the underlying gene is unknown(Fig.3).The DNA markers umc1511 and bnlg162, linked to a QTL on chromosome 4,were used to select FER-resistant lines,resulting in a 33.7%–35.2%resistance increase in a NIL carrying the homozygous resistance region compared to the background of the highly susceptible maize line Xi502[93].Ding et al.[84]identified a major QTL for FER resistance on chromosome 3 (bin 3.04), which was flanked by markers umc1025 and umc1742, facilitating markerassisted breeding for FER resistance in maize.Xia et al.[59] found that the chromosome segment substitution line CL171 harboring qFER1.03, flanked by markers Y1q25 and bumc1144, showed increased resistance to FER.Mu et al.[19] used marker-assisted selection to develop NILs carrying qRcF.verticillioides2, the major QTL for Fusarium cob rot located on chromosome 2, resulting in a 20%–35% reduction in length of infected cob.Ju et al.[17] developed NILs by introducing the locus chr1.S_6433914 from resistant line BT-1 to the background of a susceptible line N6 using markers umc1292 and umc1106,resulting in increased seed resistance to F.verticillioides.However, conducting multiple generations of backcrossing to remove background portions outside of the target locus remains a time-consuming and labor-intensive process (Fig.3).

Fig.3.Molecular design breeding for improving F.verticillioides resistance in maize.A comprehensive approach integrating transcriptomics, proteomics, metabolomics,GWAS,BSA, and linkage mapping data is essential for identifying QTL and putative genes associated with maize resistance to F.verticillioides.For marker-assisted breeding,multiple F.verticillioides-resistant QTL can be introduced into recipient lines using DNA markers linked to the QTL with multigenerational backcrossing.In the case of genetic engineering, only one generation of segregation is required after genetic transformation to obtain a transgene-free line with the desired mutation.EC, expression cassettes.

6.3.Transgenic approach and genome-editing technologies

Although many QTL associated with F.verticillioides resistance have been detected in maize, few genes have been successfully cloned and functionally validated by transgenic methods.Thus,reports of successful improvement of F.verticillioides resistance in maize by genetic engineering are limited.Public concerns about unproven health and environmental safety risks also hinder the use of transgenic maize for increasing F.verticillioides resistance[144].But the emergence of genome-editing technologies provides a new opportunity for the application of molecular design breeding,by allowing efficient induction of targeted mutations in plant genomes without introducing expression cassettes[145–147].Molecular design breeding using genome-editing technologies provides an effective and time-saving countermeasure for F.verticillioidescaused FER and FHB, especially in maize–wheat rotation systems(Fig.3).Liu et al.[25] used CRISPR/Cas9 technology to construct three homozygous lines with ZmFER1 locus mutations, and the resulting transgene-free maize plants showed increased FER resistance in three environments without evident agronomic penalty,indicating the benefits of engineering ZmFER1 for breeding FERresistant maize cultivars.A CRISPR/Cas9-caused deletion spanning the start codon of TaHRC, the homolog of ZmFER1 in wheat, conferred FHB resistance in wheat [148,149].Chen et al.[150] developed and optimized a barley stripe mosaic virus-mediated gene editing system to edit TaHRC, providing a promising approach for breeding FHB-resistant wheat without genotype limitation.

7.Future perspectives

7.1.Development of accurate and automated high-throughput phenotyping platforms

Disease severity caused by F.verticillioides, particularly for FER,is typically visually scored on scales of 1–5, 1–7, or 0–9[36,52,60,76,84,85,92,93].Such methods are prone to inconsistency due to human interpretation, limiting their accuracy for breeding and scientific research.The only reliable way for screening FER resistance is in the field,but inoculation efficiency depends on favorable environmental conditions[52].Given the large genotype and environment interactions of F.verticillioides-caused disease, field phenotypes must be conducted in multiple environments over several years and repeats, a time-, labor-, and resource-intensive process.Thus,phenotyping is the major bottleneck in identifying QTL or genes for resistance against F.verticillioides [52,151,152].To address this challenge, commercially available and automated high-throughput phenotyping platforms that measure disease symptoms caused by F.verticillioides with high precision in the field are urgently needed to accelerate the discovery of loci and genes and to characterize the resistance effects of specific genes [27,141].Laboratory techniques or seedling tests are urgently needed to test resistance shown in the field[52].The development of phenotypic identification should aim to precisely calculate the proportion of infected area to the total area using 3D image scanning and automatic calculation [153–155].Thermal-imaging and remote-sensing technologies are alternative methods for visually assessing disease indices [156–158].

7.2.Integration of multiple -omics with GWAS and linkage mapping

Plant resistance to fungal pathogens is a complex molecular process regulated by finely tuned gene networks [159].Despite the identification of numerous QTL and putative genes associated with F.verticillioides resistance using linkage mapping, GWAS,and-omics technologies,few of these genes have been functionally validated, leaving us with a poor understanding of the molecular mechanisms underlying maize defense against F.verticillioides.To address this knowledge gap,a comprehensive approach integrating transcriptomics, proteomics, metabolomics, GWAS, and linkage mapping data is desirable.By identifying genes, proteins, and metabolites with differentially expressed patterns in resistant and susceptible lines upon F.verticillioides infection, and by combining these findings with signals from GWAS and linkage mapping, the most likely putative genes responsible for F.verticillioides resistance can be identified[39].Subsequently,transgenic methods or genome-editing techniques can be employed to investigate the roles of these genes in maize resistance to F.verticillioides.Multiple-omics integration has been successfully applied to identify genes controlling crop agronomic traits [160–163],and such an approach for F.verticillioides resistance is needed.

7.3.Identification and use of desirable alleles from wild maize

The artificial domestication of maize has resulted in the loss of numerous stress-tolerance traits due to the strong selection for desirable traits that favor human consumption and adaptation to local cultivation conditions [89,139,164–166].The wild ancestor of maize,teosinte,may provide a new source of alleles to improve traits of modern maize [167,168].THP9, a major QTL for high protein content, is highly expressed in teosinte but not in the B73 inbred [169].Overexpression of THP9-teosinte in the B73 background led to an increase in seed protein content[169].Introgression of THP9-teosinte into modern maize inbred line and hybrids increased the accumulation of free amino acids and the content of seed protein without yield penalty [169].ZmMM1, a teosintederived allele of a resistance gene, confers resistance in maize to northern leaf blight, gray leaf spot, and southern corn rust [170].Several alleles derived from teosinte that confers resistance to gray leaf spot and southern leaf blight have been identified [171–173].The identification and use of desirable alleles from wild maize should be the preferred approach for maize resistance breeding to F.verticillioides.

7.4.Application of research findings from other plant species to maize

Research results from other plant species can be applied to maize.TaHRC, a gene that encodes a putative histidine-rich calcium-binding protein, was initially discovered in wheat as the main determinant of Fhb1-mediated resistance to FHB[148,149,174].Liu et al.[25]successfully engineered TaHRC homologs in maize and developed transgene-free maize plants that showed increased FER resistance without evident agronomic penalty.Given that both FER and FHB can be caused by F.verticillioides[15,25,159], engineering homologs of wheat Fhb genes such as Fhb7 [159,175], may be a feasible approach for F.verticillioides resistance breeding in maize.

7.5.Exploration of susceptibility genes for F.verticillioides

Genetic research on improving maize resistance to F.verticillioides focuses on the discovery of resistance genes.However,these resistance genes often lose their effectiveness owing to pathogen mutations that permit immune evasion [176–178].Susceptibility genes are host genes essential for pathogen virulence and pathogenicity [179,180].In contrast to that of resistance genes,resistance mediated by mutations in susceptibility genes often exhibits durability and broad-spectrum characteristics [178,180].As genome editing technologies continue to advance,the modification and editing of susceptibility genes have emerged as new approaches for increasing crop resistance [180].Wang et al.[178]recently identified and characterized a wheat receptor-like cytoplasmic kinase gene called TaPsIPK1, which was found to confer susceptibility to pandemic stripe rust, and CRISPR-Cas9 inactivation of TaPsIPK1 in wheat conferred broad-spectrum resistance against Puccinia striiformis f.sp.tritici (Pst) without affecting agronomic traits.The identification of susceptibility genes for F.verticillioides should attract more research attention.

7.6.Application of genomic selection for improving breeding of maize resistance to F.verticillioides

Research on crop disease resistance focuses primarily on identifying,locating,and selecting major resistance genes that are highly effective but susceptible to rapid changes of pathogen variants[181,182].In contrast, breeding for quantitative resistance using minor genes tends to develop more durable cultivars [181,182].Compared to marker-assisted breeding that focuses on single major resistance genes as selection targets, genomic selection breeding provides a promising alternative strategy by aggregating numerous minor-effect resistance genes [181–184].Genomic selection breeding aims to build a genetic model based on the association between molecular marker genotypes and phenotypes across the entire genome of a testing population,and then estimate breeding value and predict phenotype of the selected population with known genotypes, so as to select the best candidates as parents for the next selection cycle based on their predicted breeding values [183,184].Genomic selection breeding has become a cutting-edge technology for efficient and accurate selection of breeding populations for resistance to crop diseases such as wheat rusts and maize leaf blight [182,185,186].Given that F.verticillioides resistance is a complex quantitative trait typically governed by multiple genes with minor effects,application of genomic selection would facilitate the rapid selection of superior genotypes and accelerate the breeding cycle of maize resistance to F.verticillioides.

CRediT authorship contribution statement

Yufang Xu, Tao Li, Zhirui Zhang, Ping Lu, Ruiqi Li and Peipei Ma: wrote the manuscript and prepared the figures;Huiyong Zhang, Tao Li and Jianyu Wu: designed, supervised, reviewed,and edited the writing.

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.

Acknowledgments

We apologize to authors whose work could not be cited in this review owing to space limitations.This work was supported by the National Natural Science Foundation of China (32201787,32201793), the Innovation Special Program of Henan Agricultural University for Science and Technology(30501044),and the Special Support Fund for High-Level Talents of Henan Agricultural University (30501302).