Elevated CO2 Can Worsen Fusarium Head Blight Disease Severity in Wheat but the Fhb1 QTL Provides Reliable Disease Resistance

Fusarium head blight (FHB) is a destructive fungal disease of wheat that causes significant economic loss due to lower yields and the contamination of grain with fungal toxins (mycotoxins), particularly deoxynivalenol (DON). FHB disease spread and mycotoxin contamination has been shown to worsen at elevated CO2, therefore, it is important to identify climate-resilient FHB resistance. This work evaluates whether wheat with the Fhb1 quantitative trait locus (QTL), the most widely deployed FHB resistance locus in wheat breeding programs, provides reliable disease resistance at elevated CO2. Near-isogenic wheat lines (NILs) derived from either a highly FHB susceptible or a more FHB resistant genetic background, with or without the Fhb1 QTL, were grown in growth chambers at ambient (400 ppm) and elevated (1000 ppm) CO2 conditions. Wheat was inoculated with Fusarium graminearum and evaluated for FHB severity. At elevated CO2, the NILs derived from more FHB-resistant wheat had increased disease spread, greater pathogen biomass and mycotoxin contamination, and lower rates of DON detoxification; this was not observed in wheat from a FHB susceptible genetic background. The Fhb1 QTL was not associated with increased disease severity in wheat grown at elevated CO2 and provided reliable disease resistance.


Introduction
Fusarium head blight (FHB) is a pervasive fungal disease of cereal crops which causes significant economic damage [1,2].In north America, the primary causal agents of the disease are members of the Fusarium graminearum species complex [3,4].Fungal spores initially infiltrate through exposed florets and then hyphae quickly colonize the wheat head through the rachis [5].As the infection advances, F. graminearum produces deoxynivalenol (DON), a mycotoxin that causes significant plant cell death and leaves grains withered and contaminated [1,6].Yield and quality suffer significantly and the harvested grain, contaminated with mycotoxins, may be unsuitable for human or animal consumption [7].
The likelihood and severity of FHB infection are highly dependent on environmental conditions and field management practices.Crop residues, particularly corn stover and wheat chaff, can significantly increase the likelihood of FHB infection by providing a refuge for the pathogen to overwinter and support the formation of fungal fruiting bodies, perithecia, as the environment becomes favorable [8,9].Changes in climate are likely to shift pathogen populations and promote more severe and unpredictable disease outbreaks due to the intricate interplay between climatic, genetic, and agronomic factors [10][11][12][13].Effective control of FHB often includes multiple integrated strategies including crop rotation, tillage/debris management, fungicide application, and the selection of FHB-resistant Plants 2023, 12, 3527 2 of 15 wheat varieties [14,15].Of particular concern to breeders and growers is maintaining FHB resistance in wheat without compromising grain yield and quality; under-performing resistant varieties are likely to be abandoned [9].Furthermore, recent studies have shown that the severity of FHB infection is expected to increase with rising atmospheric carbon dioxide (CO 2 ) concentrations, prompting a search for disease-resistant cultivars with climate resilience [16,17].
Resistance to FHB is often categorized by the trait effect: Type I entails resistance to initial infection, Type II is resistance to fungal spread, and Type V is the de-toxification of mycotoxins, e.g., glycosylation of DON to DON-3-glucoside, by the host [18][19][20].However, there is currently no known trait or combination of traits which provide complete resistance to FHB.There are multiple quantitative trait loci (QTL) that provide a stable degree of resistance to the disease: Fhb1, Fhb2, Fhb5, and Fhb7; with Fhb1 being the most widely deployed.The Fhb1 QTL was first identified from the Sumai 3 wheat cultivar on the short arm of the 3B chromosome [21][22][23].Nearly all FHB resistance in north America in hard red spring wheat is derived from Sumai 3, with most breeding efforts focused on the incorporation of the Fhb1 QTL [24].While Fhb1 helps slow the spread of the pathogen, it is not involved in the formation of D3G; however, the ability to detoxify DON has also been mapped near Fhb1 on the short arm of chromosome 3B [25,26].
Functional validation of Fhb1 has not been resolved [26][27][28][29][30]. Within the Fhb1 locus, a pore-forming toxin-like gene (PFT) was identified and found to confer FHB resistance [26].This gene was predicted to encode for a chimeric lectin that may disrupt the cell membrane of fungal pathogens.However, the PFT gene was also found in FHB-susceptible wheat and may be associated with general plant defense responses to biotic and abiotic stresses [27].FHB resistance from Fhb1 may also be due to a deletion mutation in TaHRC, a histidine-rich calcium-binding-protein gene [31,32].Another candidate gene recently identified in the Fhb1 locus, TaLAC4, is associated with increased FHB resistance through the lignification and thickening of secondary cell walls.This gene was predicted to encode a laccase protein that catalyzes lignin polymerization in the wheat rachis [28].Furthermore, TaNAC032, a transcription factor within the Fhb1 locus, regulates TaLAC4 and other lignin biosynthesis genes [29].Resistance to FHB in the Sumai 3 cultivar is likely due to several general plant defense mechanisms: the induction of phenylpropanoids, the accumulation of lignin precursors, and the deposition of antifungal/antioxidant flavonoids at the cell wall all contribute to the reduction in pathogen spread and DON contamination [33].
The Sumai 3 cultivar has multiple major and minor QTLs associated with FHB resistance and extensive breeding efforts have been conducted to incorporate these resistance factors into elite wheat breeding lines; however, producing lines with acceptable crop performance has been exceedingly difficult [34][35][36].Concerningly, recent studies with 26 wheat lines found that increased FHB resistance was correlated with decreased grain protein content at elevated atmospheric CO 2 [37,38].However, Fhb1 was not associated with the disproportionate loss in grain protein content [38].Additionally, wheat grown at elevated CO 2 was found to be more susceptible to FHB spread though this effect was plant cultivar and pathogen strain specific [17].
Since the Fhb1 QTL is widely utilized for FHB Type II resistance in wheat breeding and wheat has been observed to become more susceptible to FHB spread with rising CO 2 , it is vitally important to determine whether wheat with Fhb1 retain disease resistance efficacy at elevated CO 2 .While changes in temperature and precipitation will affect the likelihood of FHB infection based on geographic location, the rising atmospheric CO 2 concentration will impact all crops globally.It is crucial for breeders to know if key loci, targeted for disease resistance, will remain reliable under future atmospheric CO 2 concentrations.
In this study, near-isogenic wheat lines (NIL) derived from either a highly FHB susceptible or a more resistant genetic background, with or without the Fhb1 QTL (Table 1), were grown in a completely random block design at ambient (400 ppm) and elevated (1000 ppm) CO 2 conditions.The near-isogenic lines were inoculated with Fusarium graminearum and evaluated for disease progression, mycotoxin contamination, and fungal biomass accumu-lation.Differences in FHB disease severity were analyzed to determine if the Fhb1 QTL provided reliable FHB resistance at elevated CO 2 .Furthermore, FHB severity was evaluated to determine whether the wheat genetic background altered Fhb1 disease resistance performance, particularly at elevated CO 2 .Furthermore, since a number of glycosyltransferases are known to be located near the Fhb1 QTL region on chromosome 3B and the fact that this proximity has led to speculation that Fhb1 was associated with DON detoxification, we also looked at the proportion of DON to D3G conversion in the NIL lines at ambient and elevated CO 2 .

Results
The Fhb1 QTL significantly reduced disease severity in wheat, maintaining a consistent percent reduction in disease spread and mycotoxin contamination regardless of genetic background or atmospheric CO 2 (Figure 1).However, the genetic background of the wheat genotypes significantly impacted disease severity at elevated CO 2 , compared with ambient levels.Genotypes HR45, HR123, and HR56, derived from a more resistant genetic background (M), had significantly greater disease severity when grown at elevated CO 2 ; the genotypes displayed greater disease spread (Figure 1a), greater relative fungal biomass accumulation (Figure 1b), and greater DON contamination (Figure 1c).
Increased pathogen biomass was observed in all but one M genotype at elevated CO 2 .The increased disease severity was already apparent only one week after point inoculation, with more rapid disease spread in the M genotypes at elevated CO 2 (Figure 2).Increased disease severity at elevated CO 2 was most pronounced in the HR123 genotype, which had severely accelerated disease spread and a more than five-fold increase in DON accumulation (Figure 1a,c).Not all M genotypes were so severely affected as neither 260-4 nor 260-2 had a significant increase in toxin accumulation at elevated CO 2 .However, both genotypes had relatively poor FHB resistance at ambient CO 2 , with 2-to 4-fold greater mycotoxin accumulation compared to the other genotypes within their Fhb1 group (M− or M+, respectively).
In comparison, the susceptible genotypes (S) had no increased susceptibility to FHB when grown at elevated CO 2 , with the sole exception of Wheaton, which had an increase in relative fungal biomass (Figure 1b).Surprisingly, the susceptible genotype Apogee, as well as its NIL A73 (Fhb1+), had less disease spread and DON accumulation at elevated CO 2 when compared with plants grown at ambient CO 2 (Figure 1a,c).However, while Apogee had reduced disease spread and DON accumulation at elevated CO 2 , it had the highest DON contamination, disease spread, and fungal biomass accumulation of any genotype tested at either ambient or elevated CO 2 .In comparison, the susceptible genotypes (S) had no increased susceptibility to F when grown at elevated CO2, with the sole exception of Wheaton, which had an incr in relative fungal biomass (Figure 1b).Surprisingly, the susceptible genotype Apoge well as its NIL A73 (Fhb1+), had less disease spread and DON accumulation at elev CO2 when compared with plants grown at ambient CO2 (Figure 1a,c).However, w Apogee had reduced disease spread and DON accumulation at elevated CO2, it had highest DON contamination, disease spread, and fungal biomass accumulation of genotype tested at either ambient or elevated CO2.
When analyzing differences between wheat groups, S genotypes were not impa by atmospheric CO2 but the M genotypes had more severe FHB infections when grow elevated CO2 (Figure 3).The M-genotypes that lacked the Fhb1 QTL had 63% more dis spread (p < 0.0001), 89% more fungal biomass (p = 0.0003), and 109% more DON accu lation (p < 0.0001) at elevated CO2 compared to ambient.While the M+ genotypes had overall disease, due to the presence of the Fhb1 QTL, they had 70% more fungal biom (p < 0.0001) and 52% more DON (p = 0.0065) at elevated CO2 (Figure 3).The Fhb1 Q provided significant FHB disease resistance regardless of the wheat genetic backgro or CO2 condition.On average, the Fhb1 QTL reduced DON by 47% (p < 0.0001), dis spread or area under the disease progression curve (AUDPC) by 45% (p < 0.0001), pathogen fungal biomass (Fg) by 46% (p < 0.0001).While the Fhb1 QTL reduced dis severity in wheat, the genetic background had a greater impact on disease outcome M genotypes had 57% less disease spread (p < 0.0001), 71% less fungal biomass (p < 0.00 and 77% less DON (p < 0.0001) compared to susceptible cultivars.Furthermore, in a mutational multivariate analysis of variance to determine what factors contributed m to the variance in disease severity, wheat genetic background accounted for 34%, w Fhb1 accounted for only 10% (Table S1).Even without the Fhb1 QTL, the M− genoty maintained a substantial degree of resistance against FHB disease spread and toxin tamination (Table S2 and Figure 3).Only at elevated CO2 was disease spread compar between S+ and M− wheat (p = 0.3791); however, both DON and fungal biomass accu lation remained significantly lower in M− genotypes (p = 0.0005 and p = 0.0058, res tively).When analyzing differences between wheat groups, S genotypes were not impacted by atmospheric CO 2 but the M genotypes had more severe FHB infections when grown at elevated CO 2 (Figure 3).The M-genotypes that lacked the Fhb1 QTL had 63% more disease spread (p < 0.0001), 89% more fungal biomass (p = 0.0003), and 109% more DON accumulation (p < 0.0001) at elevated CO 2 compared to ambient.While the M+ genotypes had less overall disease, due to the presence of the Fhb1 QTL, they had 70% more fungal biomass (p < 0.0001) and 52% more DON (p = 0.0065) at elevated CO 2 (Figure 3).The Fhb1 QTL provided significant FHB disease resistance regardless of the wheat genetic background or CO 2 condition.On average, the Fhb1 QTL reduced DON by 47% (p < 0.0001), disease spread or area under the disease progression curve (AUDPC) by 45% (p < 0.0001), and pathogen fungal biomass (Fg) by 46% (p < 0.0001).While the Fhb1 QTL reduced disease severity in wheat, the genetic background had a greater impact on disease outcome; the M genotypes had 57% less disease spread (p < 0.0001), 71% less fungal biomass (p < 0.0001), and 77% less DON (p < 0.0001) compared to susceptible cultivars.Furthermore, in a permutational multivariate analysis of variance to determine what factors contributed most to the variance in disease severity, wheat genetic background accounted for 34%, while Fhb1 accounted for only 10% (Table S1).Even without the Fhb1 QTL, the M− genotypes maintained a substantial degree of resistance against FHB disease spread and toxin contamination (Table S2 and Figure 3).Only at elevated CO 2 was disease spread comparable between S+ and M− wheat (p = 0.3791); however, both DON and fungal biomass accumulation remained significantly lower in M− genotypes (p = 0.0005 and p = 0.0058, respectively).
The percent of deoxynivalenol-3-glucoside (D3G) was evaluated to determine whether differences in glycosylation rate could account for altered FHB severity at elevated CO 2 .The proportion of DON to D3G was significantly less for most of the M genotypes grown at elevated CO 2 (Figure 4a).The decrease in percent D3G was associated with the M genetic background but was not associated with the Fhb1 QTL (Figure 4b).Interestingly, Apogee (Fhb1−) and A73 (Fhb1+) differed significantly in percent D3G (p < 0.0001), despite being derived from the same parental genetic background.The percent of deoxynivalenol-3-glucoside (D3G) was evaluated to determine whether differences in glycosylation rate could account for altered FHB severity at elevated CO2.The proportion of DON to D3G was significantly less for most of the M genotypes grown at elevated CO2 (Figure 4a).The decrease in percent D3G was associated with the M genetic background but was not associated with the Fhb1 QTL (Figure 4b).Interestingly, Apogee (Fhb1−) and A73 (Fhb1+) differed significantly in percent D3G (p < 0.0001), despite being derived from the same parental genetic background.While reduced plant detoxification of DON into D3G could account for some of the increased disease severity, percent D3G was poorly correlated with disease severity metrics in the M genotypes (Figure 5).Only in the susceptible wheat genotypes was the percentage of DON to D3G negatively correlated with disease severity.In the M genotypes, percent D3G had no significant correlation to disease spread, fungal biomass, or DON contamination.Therefore, changes in DON glycosylation, while significantly reduced in M genotypes at elevated CO 2 , was not the principal cause of increased FHB severity.While reduced plant detoxification of DON into D3G could account for som increased disease severity, percent D3G was poorly correlated with disease sever rics in the M genotypes (Figure 5).Only in the susceptible wheat genotypes was centage of DON to D3G negatively correlated with disease severity.In the M gen percent D3G had no significant correlation to disease spread, fungal biomass, o contamination.Therefore, changes in DON glycosylation, while significantly red M genotypes at elevated CO2, was not the principal cause of increased FHB sever

Discussion
Our results demonstrate that the Fhb1 QTL provided reliable FHB resistance in wheat, even at elevated CO2.Furthermore, Fhb1 provided a consistent percent reduction in disease severity regardless of the wheat genetic background: 45-47% reduction in dis-

Discussion
Our results demonstrate that the Fhb1 QTL provided reliable FHB resistance in wheat, even at elevated CO 2 .Furthermore, Fhb1 provided a consistent percent reduction in disease severity regardless of the wheat genetic background: 45-47% reduction in disease progression, DON contamination, and pathogen biomass.This was consistent with previously reported FHB disease resistance provided by the Fhb1 region in the Sumai 3/Stoa NIL lines [23].However, the overall disease severity was primarily dependent on the inherent FHB resistance, or lack thereof, in the wheat genetic background.While Fhb1 reduced FHB spread in S+ genotypes compared with the highly susceptible S− genotypes, mycotoxin contamination was still substantial in all wheat from a susceptible genetic background.
M genotypes from the Sumai 3/Stoa RIL 63-4//MN97448 pedigree suffered increased disease susceptibility to FHB spread at elevated CO 2 unlike wheat from susceptible genetic backgrounds.However, the Fhb1 QTL was not associated with increased relative disease susceptibility at elevated CO 2 as M-genotypes, which lacked Fhb1, had more severe FHB infections and contamination when grown at elevated CO 2 .FHB disease resistance in wheat has been associated with the loss of grain protein and mineral content at elevated CO 2 [37,38].The accessibility of host nutrients and the availability of nitrogen can substantially influence the expression of F. graminearum genes associated with virulence and pathogenicity [39,40].Fungal pathogens often have a shared promotor region that links virulence-associated genes with major nitrogen regulatory transcription factors [41].The F. graminearum strain used in this study (9F1) has been previously observed to increase mycotoxin biosynthesis in response to the loss of wheat nutritional content [42].Furthermore, the 9F1 strain has caused differential disease severity in wheat depending on the CO 2 concentration and the genetic background of the host, though that study employed a whole-head inoculation method rather than the single floret inoculation method used in this study [17].As we observed in the M genotypes, F. graminearum had greater disease spread, greater fungal biomass, and more DON accumulation in the moderately resistant wheat cultivar Alsen-harboring Fhb1 and derived from Sumai 3-when it was grown at elevated CO 2 .
In the current study, the most significant contributing factor to overall disease resistance was the wheat genetic background.Plant defense against pathogens can involve physiological, phytochemical, and compositional adaptations as well as a plethora of kinases, glycosyltransferases, antimicrobial peptides, and apoplastic proteases, though these defenses are often countered by antagonistic microbial virulence factors [18,[43][44][45].While the Fhb1 QTL from Sumai 3 is the most widely studied and deployed, there are numerous other wheat QTLs associated with FHB resistance; most are considered minor contributors to plant defense but few have been validated and incorporated into breeding programs [46,47].However, the stacking or combination of multiple resistance alleles can significantly enhance wheat defenses against head blight [48].This is evidenced in our results (Figure 3) where the M− genotypes still had strong disease resistance even though the expected main source of resistance, Fhb1, was bred out of the NILs.Combining the highly effective Fhb1 QTL with the putative resistance traits within the M genetic background resulted in the lowest disease severity, even with worsening FHB infection and DON contamination at elevated CO 2 .We did find a significant Fhb1*genetic background interaction (p < 0.0001; Table S1), indicating that Fhb1 was more effective in more resistant genetic backgrounds.This suggests synergistic interactions between Fhb1 and the yet unidentified resistance factors within the M genotypes.However, fully elucidating these interactions is beyond the scope of the current work.
Identifying specific resistance genes and evaluating their functionality is an incredibly arduous undertaking.Furthermore, unwanted or deleterious genes often accompany the desired resistance trait during plant breeding, resulting in linkage drag or the reduction in plant fitness from the genetic stowaways [36].While Fhb1 is not associated with an altered plant response or increased FHB susceptibility at elevated CO 2 , the M genetic background harbors currently unidentified detrimental trait/s that cause these effects.
From our results, we found a consistent reduction in type II and type V resistance in M genotypes at elevated CO 2 .
The Fhb1 QTL only provides Type II resistance or the inhibition of disease spread and severity and does not affect the likelihood of infection [21,23].Wheat is also capable of converting the phytotoxic DON into D3G, detoxifying the trichothecene through the addition of glucose-a form of type V resistance [49].The modification of bioactive compounds by glycosylation, namely the conjugation of a sugar molecule to a compound via a glycosyltransferase enzyme, is a common feature of plant defense used to alter toxin bioavailability to hosts or pathogens [50,51].The wheat genome contains hundreds of glycosyltransferases located among all chromosomes and across all three sub-genomes (A, B, and D) that strongly respond to pathogen infection.In one study, 59% of these glycosyltransferases were upregulated two days after infection by a DON-producing F. graminearum strain compared with control wheat [52].However, the ability of wheat to convert DON to D3G was only helpful in reducing disease severity for the susceptible genotypes where the percentage of DON converted was negatively correlated with disease spread, fungal biomass, and total DON accumulation (Figure 5).We had hypothesized that the glycosylation of DON and, therefore, the inactivation of its phytotoxicity, would be an important factor in plant FHB resistance but the conversion of DON to D3G appears to provide minimal protection compared to other resistance factors found within the Sumai 3 genetic background.Our results suggest that it may be advantageous in terms of FHB resistance to seek wheat breeding strategies that incorporate non-native or novel DON detoxifying mechanisms-such as Fhb7 [53]-that could provide more protection than native wheat glycosyltransferases.
The waning FHB resistance of M genotypes with rising CO 2 is concerning for the future of FHB mitigation in wheat.Utilizing FHB-resistant wheat cultivars is an essential component in an effective integrated management strategy to reduce FHB damage [54].Further breeding efforts are needed to avoid diminishing FHB resistance, coupled with reduced grain nutritional quality, as atmospheric CO 2 concentrations increase.But despite worsening disease severity in the more FHB-resistant wheat, they remained the most resilient to mycotoxin contamination with rising CO 2 (Figure 3).Therefore, the use of FHB resistance factors from Sumai 3, particularly Fhb1, should still be utilized in wheat breeding programs for reducing mycotoxin contamination and disease damage.Efforts are ongoing to identify the cause of the detrimental CO 2 response in the M genotypes.It is vitally important to identify climate-resilient and disease-resistant wheat cultivars to ensure future food safety and security.

Fhb1 Near-Isogenic Lines
Two sets of NILs with varying inherent FHB resistance were used to determine the impact of elevated CO 2 on Fhb1 disease resistance efficacy.Details on the backcrossing and characterization of the susceptible NILs were reported previously [38].Briefly, the first set of NILs were produced from highly FHB susceptible cultivars (S), Norm [55], Wheaton [56], and Apogee [57,58].Apogee, Norm, and Wheaton are often used as susceptible checks in FHB research.For this manuscript, genotypes derived from these highly FHB susceptible cultivars are designated as genetic background 'S'.Sumai 3 was used as the donor of Fhb1 to generate near-isogenic lines in the three susceptible genotypes using markerassisted backcrossing.The Xgwm493 molecular marker [59] was used to select for Fhb1 in four generations with each respective genotype serving as a recurrent parent.The susceptible BC 4 F 3 Fhb1 near-isogenic lines are designated N1 (Norm near-isogenic line), W4 (Wheaton near-isogenic line, and A73 (Apogee near-isogenic line) and are more than 97% homozygous with the recurrent parent genome.NIL genotypes from the susceptible check genetic background are referred to as S− (for those without the Fhb1 QTL) and S+ (for those with the Fhb1 QTL).The second set of NILs was developed using a moderately susceptible spring wheat background that has been used for the fine mapping of Fhb1 [60].The choice of the FHB moderately susceptible background was to allow for the characterization of genotypes that lacked the FHB resistance provided by Fhb1, i.e., mapping the genomic region harboring Fhb1.For the purpose of clarity, this group of NILs is designated genetic background 'M'.Genotypes that possess a functional Fhb1 QTL were classified as moderately resistant (M+) while those without Fhb1 were classified as moderately susceptible (M−).The NIL set was derived from a single F7 plant that was heterozygous for Fhb1 and all NILs have the pedigree Sumai 3/Stoa RIL 63-4//MN97448.The genotypes are designated 260-2, HR 45, HR 56, HR 58, HR 123, and 260-4.All NIL lines were examined previously for agronomic performance at elevated CO 2 [38].

Wheat Growing Conditions
To evaluate whether FHB resistance from the Fhb1 QTL is diminished at elevated CO 2 , the NIL wheat genotypes (Table 1) were grown in PGR15 environmentally controlled growth chambers capable of CO 2 fumigation (Controlled Environments INC., Winnipeg, MB, Canada).Growing conditions were similar to those reported previously [38].Briefly, plants were grown in six chambers, three set to ambient CO 2 (420 ± 10 ppm, a[CO 2 ]) and three set to elevated CO 2 (1000 ± 10 ppm [CO 2 ], (e[CO 2 ]).The elevated CO 2 treatment was chosen based on the predicted atmospheric CO 2 concentration at the end of the century in the highest emissions scenario (SSPS-8.5)[61].For each chamber, wheat genotypes were grown in triplicate 20 × 15-cm plastic pots filled with approximately 4 L of SunGrow Horticulture potting mix (Agawam, MA, USA), with 5 plants per pot.The experiment was performed twice, evaluating half of the NILs at a time.The wheat was grown at 25/23 • C (day/night) and 50-60% relative humidity.The chamber light source was from a mix of incandescent and fluorescent lights set to a 14 h photoperiod (550 µmol m −2 s −1 photosynthetic photon flux density).Plants were well watered, typically daily, and received additional biweekly fertilization with Peters 20-20-20 nutrient supplement (The Scotts Company, Marysville, OH, USA) until anthesis.

Disease Assays
To evaluate Fhb1 disease resilience at elevated CO 2 , the F. graminearum isolate 9F1 (NRRL 37676) was used to inoculate wheat cultivars (Table 1) at anthesis (flowering).The 9F1 strain is a DON producer originally isolated from wheat in the Netherlands.The preparation of media, pathogen culturing, and wheat inoculations were performed according to previously reported methods [42].Briefly, fungal isolates were grown on V8 agar plates for 7 d before an agar plug was transferred into 20 mL of sterile mung bean broth.Cultures were grown for 48 h, at 28 • C, and 160 rpm under dark conditions in a New Brunswick Innova 44 incubator shaker (Eppendorf, Hauppauge, NY, USA) to promote conidia formation.The culture was centrifuged and the mung bean supernatant was discarded.Afterwards, the conidia pellet was resuspended in 0.04% Tween 20 (Thermo Fisher Scientific, Waltham, MA, USA) to form a 1 × 10 5 mL −1 conidia suspension.The suspension was used for wheat floret inoculations immediately after preparation.
Fifteen wheat heads per line, per chamber, were tagged and inoculated at flowering with 10 µL of the conidial suspension into single florets.Immediately after inoculation, plastic bags were placed over the wheat heads to maintain a high humidity environment.The plastic bags were supported by a central bamboo stake to prevent lodging or severe stem bending and the stake(s) and bag(s) were removed after 3 d.Disease progression and the area under the disease progression curve (AUDPC) was determined by visual assessment of the number of diseased florets (identified as bleached or necrotic plant tissues) on days 7, 10, 14, 17, and 21 after inoculation.At day 21, the infected wheat heads were collected and stored at −80 • C before being lyophilized and ground for mycotoxin and pathogen biomass analysis.

Mycotoxin Analyses
To determine how Fhb1 impacted mycotoxin accumulation in wheat heads at ambient and elevated CO 2 , we measured both deoxynivalenol (DON) and DON-3-glucoside (D3G) in infected wheat heads collected 21d after inoculation.Briefly, 0.5 g of ground infected wheat heads was extracted with 10 mL of acetonitrile/water (86:14, vol/vol).An aliquot of the extract was analyzed for glycosylated deoxynivalenol (deoxynivalenol-3-glucoside; D3G) content using liquid chromatography/mass spectrometry coupled with electrospray ionization tandem mass spectroscopy (LC-ESI-MS/MS) on an Agilent 1100 HPLC (Agilent Technologies, Inc., San Clara, CA, USA) linked with a SCIEX 3500 TripleQuad MS (AB SCIEX LLC, Framingham, MA USA).Samples were separated on a Phenomenex (Torrance, CA, USA) Kinetex C18 column (150 mm length, 4.6 mm diameter, 2.6 µm particle size) following a previously reported methodology [19].A separate 5 mL aliquot of each acetonitrile/water extract was cleaned with a MycoSep 225 column (Romer Labs, Union, MO, USA).To prepare DON trimethylsilyl (TMS) derivatives, a 2 mL aliquot of the purified extract was dried and reacted with 100 µL of a 100:1 mixture of N-methylsilylimadazole/trimethylchlorosilane (Sigma-Aldrich, St. Louis, MO, USA).After 30 min, 1 mL of water and 900 µL isooctane were added and the mixture was gently vortexed.The top isooctane fraction was analyzed with GC-MS on an Agilent 7890 gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) fitted with a HP-5MS column (30 m, 0.25 mm, 0.25 µm) and a 5977 mass detector using selective ion monitoring.TMS derivatives of purified DON (0.3125 to 80 µg) were similarly prepared to construct a standard curve.

Estimation of Host and Pathogen Biomass
The assessment of relative fungal biomass in diseased wheat heads was determined using the ratio of Fg DNA to wheat DNA via quantitative polymerase chain reaction (qPCR) according to the methods reported previously [42].Three technical replications were performed per sample.The primers and probes used for qPCR can be found in Table 2.The approximate fungal biomass (relative Fg DNA/wheat DNA) was calculated by dividing the geometric mean of initial DNA concentration (N 0 ) from the Fusarium probes by the geometric mean of N 0 from the wheat probes.Finally, the amount of DON production per unit biomass was estimated by dividing the µg g −1 DON by the relative pathogen biomass, as quantified by qPCR.

Statistical Analyses
Results were evaluated using a non-parametric Van der Waerden test (α = 0.05) to determine significant differences due to elevated CO 2 .Subsequently, a post hoc analysis was performed evaluating group differences via a multiple pairwise comparison using a non-parametric Wilcoxon/Kruskal-Wallis test (α = 0.05).Pearson correlations between independent variables were determined by a multivariate analysis, α = 0.05 (JMP V15.0).A principal component analysis was performed in JMP V15.0.Additionally, a permutational multivariate analysis of variance was performed in R 4.2.1 (adonis2; vegan package) to analyze sources of variation in disease severity treatments.Details on pairwise comparisons can be found within the table and figure legends.

Conclusions
Fhb1 QTL was not associated with increased disease susceptibility at elevated CO 2 and provided reliable FHB resistance.However, wheat from the M genetic background suffered greater FHB disease progression, DON contamination, and pathogen biomass accumulation at elevated CO 2 .Furthermore, there was a reduction in the detoxification of DON into D3G but this alone did not account for the increased disease severity.The wheat NILs derived from susceptible genetic backgrounds did not have greater disease severity or altered rates of DON detoxification at elevated CO 2 .Inserting Fhb1 into susceptible cultivars can significantly reduce disease, but alone, the QTL does not provide robust protection; a combination of resistance factors will likely be required to strongly suppress FHB infection.The use of FHB resistant cultivars and traits, including those derived from Sumai 3, will likely remain one of the most effective strategies to reduce disease and mycotoxin contamination in wheat.However, it is essential to develop climate resilient disease resistant wheat to ensure global food safety and security.

Plants 2023 ,Figure 1 .
Figure 1.Disease severity of Fusarium graminearum infected wheat heads as defined by(a) t under the disease progression curve (AUDPC), (b) deoxynivalenol (DON) contamination, relative fungal biomass (Fg).Genotypes were derived from either a susceptible check (S) or resistant genetic background (M) and each either possessed the Fhb1 QTL, as indicated by a lacked the Fhb1 QTL, as indicated by a (−).Wheat was grown at either ambient (a[CO2]) or e carbon dioxide concentrations (e[CO2]).Error bars represent the standard error.Symbols ( and ***) denote a statistically significant effect (p < 0.1, p < 0.05, p < 0.01, and p < 0.0001, respe of elevated CO2 on AUDPC, DON, or Fg within a genotype, as determined by a non-paramet der Waerden test (AUDPC: n = 45 and DON and Fg: n = 15).

Figure 1 .
Figure 1.Disease severity of Fusarium graminearum infected wheat heads as defined by(a) the area under the disease progression curve (AUDPC), (b) deoxynivalenol (DON) contamination, and (c) relative fungal biomass (Fg).Genotypes were derived from either a susceptible check (S) or a more resistant genetic background (M) and each either possessed the Fhb1 QTL, as indicated by a (+), or lacked the Fhb1 QTL, as indicated by a (−).Wheat was grown at either ambient (a[CO 2 ]) or elevated carbon dioxide concentrations (e[CO 2 ]).Error bars represent the standard error.Symbols ( †, *, **, and ***) denote a statistically significant effect (p < 0.1, p < 0.05, p < 0.01, and p < 0.0001, respectively) of elevated CO 2 on AUDPC, DON, or Fg within a genotype, as determined by a non-parametric Van der Waerden test (AUDPC: n = 45 and DON and Fg: n = 15).
greater mycotoxin accumulation compared to the other genotypes within their Fhb1 gr (M− or M+, respectively).

Figure 2 .
Figure 2. Wheat genotypes HR123 (M−) and HR 56 (M+) at one week post single floret inocula of Fusarium graminearum.Both cultivars exhibit more severe FHB disease spread at elevated CO

Figure 2 .
Figure 2. Wheat genotypes HR123 (M−) and HR 56 (M+) at one week post single floret inoculation of Fusarium graminearum.Both cultivars exhibit more severe FHB disease spread at elevated CO 2 .

16 Figure 3 .
Figure 3. Disease severity in various wheat groups, as defined by (a) the area under the disease progression curve (AUDPC), (b) the relative fungal biomass in infected wheat heads (Fg), and (c) the accumulation of deoxynivalenol (DON) in wheat heads at either ambient (a[CO2]) or elevated carbon dioxide concentrations (e[CO2]).Error bars represent standard error.Different letters denote statistically significant differences as determined by a non-parametric Wilcoxon/Kruskal-Wallis for multiple pairwise comparisons (AUDPC: n = 135; DON and Fg: n = 45).

Figure 3 .
Figure 3. Disease severity in various wheat groups, as defined by (a) the area under the disease progression curve (AUDPC), (b) the relative fungal biomass in infected wheat heads (Fg), and (c) the accumulation of deoxynivalenol (DON) in wheat heads at either ambient (a[CO 2 ]) or elevated carbon dioxide concentrations (e[CO 2 ]).Error bars represent standard error.Different letters denote statistically significant differences as determined by a non-parametric Wilcoxon/Kruskal-Wallis for multiple pairwise comparisons (AUDPC: n = 135; DON and Fg: n = 45).

Table 1 .
Breeding pedigrees for wheat cultivars in the current study.Wheat genotypes are either derived from backcrossing Fhb1 from Sumai 3 into susceptible wheat cultivars (S background) or by identification of near-isogenic lines (NILs) in a moderately susceptible (M) background.Lines containing the Fhb1 QTL are indicated with a "+" and those without Fhb1 with a "−".

Table 2 .
Primer and probe sequences used for quantitative polymerase chain reaction (qPCR) amplification.Three sets of primers and matching probes were used to quantify relative fungal or wheat biomass.