5.2. Screening for Fusarium and Trichothecene Resistance in Cereals
Type I and II FHB disease resistances are the best characterized and the easiest to evaluate. Type I resistance quantifies resistance to initial infection expressed as a percentage of diseased spikes (disease incidence), whereas Type II resistance measures resistance to fungal spread within the spike mostly given as a percentage of disease spikelets within infected spikes (head severity). Often an ‘FHB index’ is calculated by multiplying the disease incidence × head severity/100 [120
]. Under suitable environmental conditions where the temperature and humidity are favorable for disease, numerous additional factors can unduly influence the evaluation of incidence, for example some heads within a plant or plot may escape inoculation due to their position within the canopy or their developmental stage. Type II resistance can be more reliably determined and is measured by point inoculation, where spores are injected/pipetted into an individual floret or spike, and disease spread is reported as the number of diseased spikelets within 18–20 days after inoculation. Shaner [298
] provides a valuable discussion on screening methods to evaluate different types of resistance and describes therein some of the challenges associated with screening for Type III to V resistance.
Detailed descriptions of inoculation methods are available from different sources [299
]. Alternative methods to evaluate disease resistance have also been described that are designed to save time and/or enable high-throughput analyses (for example, [304
]). Most of these involve analysis of different tissues of cereals such as seedlings or roots; for an excellent review of different inoculation methods in different tissues see Miedaner [306
]. While the pathology and mechanisms of resistance in different organs are unlike those of FHB, these assays can sometimes serve as a valuable tool to test a hypothesis and/or identify candidate genes involved in resistance. For example, Wang et al. [307
] investigated molecular aspects of resistance to FRR and determined that similar defense response genes are activated in the roots as observed in the spikes following FHB infection. Fusarium seedling blight, root and crown rots are different diseases of cereals caused by trichothecene-producing Fusarium
species, and while they tend to affect plant survival and yield, they do not typically result in toxin contamination of grain. However, in the case of FCR, the infection can move up into the spike and result in DON contamination of seeds [23
]. With the exception of FCR [308
], these diseases do not have notable economic significance.
When assessing FHB, in addition to providing suitable environments for fungal infection, such as mist-irrigation in greenhouse and field settings to increase humidity [301
], the influence of the growth stage is critical in disease assessment as this can influence interpretation of results. Spikelets inoculated prior to anthesis typically do not develop FHB symptoms. In barley, anthesis occurs within the boot, and as a result the spikes are not threatened by the disease pre-anthesis. When comparing inoculation in barley at different developmental stages after heading, McCallum and Tekauz observed no differences in disease response. Although, later infections may be symptomless while still accumulating mycotoxin [309
]. In barley there is a low correlation between Fusarium
-damaged kernel (FDK) and DON; FDK/DON correlations are even lower than that between FHB and DON. Thus, a grain sample that seems acceptable based on color, plumpness, and protein can carry high levels of DON [310
]. In wheat the disease is most severe when spikes are inoculated at anthesis. Del Ponte et al. [312
] compared disease severity, FDK and DON accumulation in the wheat cultivar Norm at six reproductive stages. Visible symptoms were most severe when inoculated at anthesis and decreased with inoculations at different stages of kernel development until the early dough stage. Meanwhile, DON occurred at all stages assessed, including the hard dough stage where kernel weight was unaffected and FDK symptoms were limited. These studies point to the importance of performing DON content evaluation in both wheat and barley.
Measurements of trichothecene accumulation can be carried out by immunodetection, such as enzyme-linked immunosorbent assays (ELISAs) that are commonly employed for high-throughput detection of DON and/or related mycotoxins. These methods are limited in that they provide information on individual toxins rather than a complete toxin profile, whereas mass spectrometry-based methods enable detection and quantification of all known toxins within an extract. More details on trichothecene quantification methodologies can be ascertained from the following articles: [26
5.4. Breeding for Low Trichothecence Content in Wheat and Barley Grains
Growing resistant cultivars is pivotal in Fusarium
disease control and for the prevention of mycotoxin contamination, but resistance breeding is complicated by the quantitative nature of the trait involving multiple genes with small to medium effects and the interaction with environmental conditions [321
]. Basically, resistance breeding relies on the available variation for the trait of interest and methods/tools to reliably measure or predict resistance levels and trichothecene contents in a breeding program. In both wheat and barley, genetic variation for FHB resistance is broad comprising ‘native’ and ‘exotic’ resistance sources. The difficulty for a breeder is to combine high yield and quality performance with resistance to other relevant diseases and pests including FHB. As a consequence, large numbers of breeding lines need to be screened in FHB disease nurseries to select superior lines favoring FHB disease assessments on the plants which are technically easier, faster and cheaper compared to direct mycotoxin quantifications.
In wheat, the relationship between visual disease evaluations in the field or on the harvested grains and DON content was investigated in a broad meta-analysis by Paul and coworkers [322
]. They analyzed 163 studies resulting in overall positive and significant correlations, with FDK showing the strongest association with DON content (r = 0.73) followed by disease index (r = 0.62), severity (r = 0.53) and incidence (r = 0.52). The role of resistance in toxin control was also highlighted in a review [323
] concluding similarly that breeding new cultivars with increased FHB resistance will result in reduced DON contamination, but moreover, reduces simultaneous levels of less prevalent and less frequently measured trichothecenes. For instance, consider the following field experiment where 190 winter wheat lines were inoculated with DON-producing F. graminearum
and T-2/HT-2 toxin-producing F. sporotrichioides
in separate trials and evaluated for symptom severity on the spikes and grains as well as DON and T-2/HT-2 toxin content. In this study, resistance measures correlated highly with DON and T-2/HT-2 toxin content within the trials and also with F. graminearum
and F. sporotrichioides
disease parameters across trials, but most noteworthy DON and T-2/HT-2 toxin contents were also associated (r = 0.80) demonstrating that indirect selection for low T-2/HT-2 toxin contents is feasible [323
] and underlining the non-species-specific FHB resistance [324
]. The effect of FHB resistance breeding in wheat on DON and its masked form D3G was discussed in a review by Lemmens et al. [325
]. Several independent experiments revealed highly significant relationships between FHB symptoms on wheat heads, DON and D3G content indicating that selection of improved lines based on FHB symptoms or DON results in a concomitant reduction in D3G content [325
Unlike wheat, a robust correlation between FDK and DON has not been reported in barley. However, Tucker et al. [328
] recently reported a strong correlation between D3G content with DON and 3-ADON. These observations indicate that the selection of reduced DON content for development of improved barley breeding lines would likely result in lower D3G content.
QTL analysis unravels the genetic architecture of Fusarium
/trichothecene resistance in a specific population in terms of the number and effect of QTL and elucidates the genetic basis of the trait associations. Twenty years ago, the first FHB resistance QTL were identified in wheat [329
] and since then numerous QTL have been reported and summarized by Buerstmayr et al. [321
] and by Buerstmayr (2019, review in preparation). The genetic basis of FHB resistance has also been studied through QTL analysis in barley and was reviewed by Kolb et al., 2001 [330
], and later by Massman et al., 2011 [331
Of particular interest in the context of this review are QTL (or genes) that confer resistance to trichothecenes and/or were found associated with reduced trichothecenes contents in the harvested grains of wheat and barley. Among the numerous FHB mapping studies in wheat, only 25 included DON measurements and provided further parameters for FHB resistance, for instance, the percentage of infected heads or spikelets, FDK, FHB severity or area under the disease progress curve (AUDPC) as a measure for overall field resistance or FHB index. These 25 studies identified 63 QTL linked with reduced DON contents of which 54 coincided with QTL for FHB parameters evaluated on the heads or grains, demonstrating a common genetic basis of FHB and DON resistance. Merely 9 QTL were detected exclusively for DON. Table S1
provides the complete list of wheat resistance QTL from experiments conducting DON measurements, whereas Table 5
excerpts major ‘DON resistance determinants’ including associations with loci controlling phenological and morphological traits.
The most prominent FHB resistance QTL in wheat, Fhb1
, derived from Chinese germplasm, confers resistance to fungal spreading and reduces DON contents in grains as confirmed by many independent studies (Table 5
governs ‘DON resistance in the narrow sense’; Lemmens et al. [336
] showed by infiltrating pure toxin solution in wheat florets, that this locus enhances the hosts’ ability to detoxify DON. Lines carrying Fhb1
conjugated almost all the applied DON into D3G whereas from those lacking Fhb1
a high percentage of the infiltered DON was recovered. Thus, it was proposed that Fhb1
either encodes or regulates a uridine diphosphate-glycosyltransferase (UGT). Fine-mapping revealed the complete contig sequence of the resistance locus, yet, no UGT was annotated [348
]. A pore-forming toxin-like gene has been isolated and is responsible for resistance to fungal spreading at the locus [349
]; however, recently two studies rebutted this finding, both identifying a critical deletion in the same gene encoding a histidine-rich calcium-binding protein as the causative mutation at Fhb1
. Su et al. [350
] concluded that the Fhb1
-derived resistance is the result of a loss-of-function mutation, whereas Li et al. [351
] demonstrated that the same deletion generates a gain of function [352
]. Whether the histidine-rich calcium-binding protein also controls DON detoxification is yet unclear.
, other large effect FHB resistance QTL, Qfhs.nau-2DL
], and Fhb7AC
] also lead to reduced disease severity and reduced toxin content simultaneously. Morphological and phenological traits such as plant height, spike/flower morphology, and heading date affect fungal infection and spreading of the disease favoring taller genotypes with a lax spike type and high anther extrusion after anthesis. Furthermore differences in heading dates often result in disease escape predominantly because of environmental conditions [321
]. Key determinants for these traits colocalize with QTL for FHB resistance and DON content, e.g., the height controlling loci Rht-B1
], the q
locus controlling spike type [346
] and the Ppd-D1
] and Vrn-A1
] loci affect heading date and plant height. Common factors for anther extrusion and DON content were detected on chromosomes 2DLc [345
] and at the Rht-B1
]. Whether all these relationships are due to pleiotropic effects or are caused by linked genes remains unknown. Few QTL were identified exclusively for DON which may indicate that plant genes exist which have a function in detoxification, but are not associated with FHB resistance in the narrow sense [354
], although overlapping with QTL for plant height [355
], anther extrusion [354
] and heading date [358
] was found for three of these loci. Also, a recent meta-QTL analysis of FHB resistance in wheat positioned the 63 QTL for DON resistance within 40 of the 65 generated meta-QTL mainly overlapping with Type I and II resistance QTL [359
To date, most genetically mapped resistance QTL has no biological function assigned, but this has not affected their deployment in resistance breeding using phenotypic and/or genotypic selection. Molecular mapping studies yield markers linked with resistance QTL and enable selection of improved individuals based on genetic fingerprints. The so-called marker-assisted selection (MAS) is successfully applied to introgress well-characterized, large-effect QTL reducing FHB severities and DON contents [353
]. Miedaner et al. [353
] reported that marker-selected lines carrying the major resistance QTL Fhb1
reduced average DON content by 59 and 43% compared to their sister lines lacking the resistance alleles, and DON content was lowered by 79% in lines with combined resistance QTL alleles. However, many of the genes contributing to resistance have small effects, especially those derived from non-Chinese sources, unsuitable to track with few markers. The common practice to pyramid these by phenotypic selection can nowadays be accelerated by genomic selection, thereby estimating genome-wide marker effects in a phenotyped training population and predict genomic estimated breeding values of non-phenotyped individuals in a selection population [363
]. The applicability of genomic selection for FHB and DON resistance breeding has been demonstrated in several studies [363
]. Regarding DON content, both genomic-estimated breeding values and phenotypic selection would select superior lines that are encouraging from a breeding perspective, as genomic selection can save time and resources associated with phenotyping [363
Relatively speaking, fewer FHB and DON QTL studies have been carried out in barley compared to wheat, with the majority of them being conducted in six-row barley. It was identified that resistance to FHB and DON accumulation are controlled by many QTL located on all seven barley chromosomes [330
]. The detected QTL are often minor, environmentally specific, and associated with phenological and morphological traits. Taller stature, late heading, row-type, lax and nodding spike are commonly associated with FHB resistance [311
]. Such unfavorable associations complicate introgression of FHB resistance into elite germplasm and are limiting the use of MAS in barley breeding programs.
Several examples of QTL from experiments that include DON measurements as well as associations with loci controlling phenological and morphological traits are presented in Table 6
. The coincidence of DON QTL with FHB severity QTL mainly translates into lower DON accumulation being associated with reduced disease severity. However, there are some reports where lower DON QTL coincided with increased FHB severity QTL [366
]. In addition, the case of QTL identified exclusively for DON and not coincident with FHB severity indicates that DON accumulation may not always be a pleiotropic effect of FHB QTL [366
] and different genes might be responsible for it. The most commonly identified QTL for reducing DON concentration were detected on the chromosome 2H with the resistance allele contributed by resistant sources such as Chevron [367
], CI 4196 [369
], and Fredrickson [370
]. These QTL have also been associated with a major heading date QTL and spike morphology controlled by VRS1
. The recent work by Huang et al. [371
] re-emphasizes the relationship between morphological characteristics and the FHB response and advocates that plant architecture and inflorescence traits must be absolutely considered when breeding barley for FHB resistance.
In 2013, an extensive germplasm screening was carried out by Huang et al. [376
] on a global collection of 23,255 wild and cultivated accessions. This analysis identified only 78 FHB resistant or moderately resistant sources. These genotypes were further haplotyped with markers associated with consistently detected FHB QTL, located on chromosome 2H and 6H, from different resistant sources (Chevron, CI 4196, Fredrickson, etc.) by previous studies [331
]. It was identified that the most common haplotype on all four QTL regions was one of the resistant source Chevron. Moreover, few other sources (cultivated or wild) with potentially novel alleles were identified based on their distinct haplotype patterns at these four QTL. That said, further mapping studies are required, which should also include DON measurements, before they can be deployed in the barley breeding programs.
Massman et al. [331
] conducted a genome-wide association study using 768 advanced barley breeding lines and identified four QTL for FHB resistance and eight QTL for DON accumulation. DON concentration QTL was identified on every chromosome with four of them being in the same regions of chromosome 1H, 2H and 4H as those previously identified in bi-parental mapping populations [367
]. Notably, a good portion of the QTL identified in this study are in many cases free of undesirable linkages makes them excellent candidates for MAS. Similarly, Mamo and Steffenson [377
], when assessing a diverse collection of barley landraces from Ethiopia and Eritrea by genome-wide association, also found that the FHB resistance and DON concentration QTL identified were not significantly associated with heading date or plant height. Despite these new findings that could potentially enable MAS in barley, the authors [331
] are proponents of the use of a more discursive marker selection strategy, such as genomic selection, as a way to amass the beneficial alleles in the barley breeding populations.
Sallam and Smith [378
] compared genomic and phenotypic selection in five sets of spring six-row barley breeding lines for yield, FHB severity and DON concentration, and concluded that the use of genomic selection for these traits in barley breeding should result in gains similar to the ones obtained using phenotyping selection but in a shorter time frame and with a lower cost. Selection gains when using genomic selection for FHB and DON are also reported in two-row barley (James Tucker, Agriculture, and Agri-Food Canada, Brandon, personal communications). Moreover, Tiede and Smith [379
] recently provided additional evidence for the effectiveness of genomic selection in six-row barley by demonstrating significant gains for these two unfavorably correlated quantitative traits, yield, and DON.
In addition, in vitro selection methods, which employ selection pressure for FHB and/or trichothecene resistant hexaploid wheat [380
] and two-row [382
] and six-row barley [384
] have been established. With improved methodologies for green plantlet regeneration [385
], this technology can now be employed on F1
hybrids, making this an attractive avenue to generate doubled haploid populations for FHB and trichothecene resistance breeding, which can also be combined with molecular genotyping approaches.
Resistance breeding programs need to find the optimal strategy within these complementary approaches according to their needs; ultimately, FHB resistant and productive cultivars are a sustainable, environmentally friendly, and economic way towards increasing food and feed safety and security.