Final Report for GNE14-092
From August 2014 to December 2015, we worked to develop an evaluation and prediction system of Gibberella seedling blight (GSB)resistance among New York grain maize hybrids. We started bydeveloping a fungal inoculation and plant growth protocol in a controlled growth chamber environment. Using this protocol, we evaluated fungus- and mock-inoculated maize seedlings for a suite of quantitative traits witha few diverse genetic inbred maize lines and a field-collected F. graminearumisolate. These traits included fungus-induced changes in seedling morphology, microscopic quantification of fungal tissue, constitutive and fungus-induced content of plant defense-related compounds, and expression of fungus-infection and plant defense-related genes. From this pilot experiment, we determine thatfungus-induced changes in maize seedling morphology are the most reliable traits that can be measured repeatedly. Therefore, we repeated these morphological measurements onnine commercial organic maize hybrids of diverse maturation time and seedling vigor and fourF. graminearumisolates collected in New York that showed a broad range of aggressiveness in a previous study.
Finally, we investigated the relationship between these fungus-induced traits and overall GSB disease severityunderour controlled growth conditions. We found that fungus-induced seedling elongation is significantly correlated with lodging rate across host and fungal genotypes, and hence could be used to evaluateGSB resistance in other maize hybrids and have the potential to predict GSB severity underfield conditions. In conclusion, we found that F. graminearum infection can induce seedling shoot elongation androot length reduction. Remarkably, seedling elongation as early as3 days post-inoculation is significantly correlated with lodging rate measured 11 days later (2 weeks post-inoculation). Hence, this easy-to-measure trait can potentially serve as an early indicator of F. graminearum seedling blight resistance in commercial maize varieties.
Fusarium graminearum is one of the most widespread fungal pathogens associated with grain corn production in New York State, causing tissue rot in roots, stalks, and ears. Depending on the timing and site of infection, F. graminearum can result in Gibberella seedling blight (GSB), Gibberella stalk rots and root rots, and pre- and post-harvest ear rot. In addition to causing direct lossesthrough tissue rot, F. graminearumpresents a severe public health risk by producing mycotoxins, including zearalenone, nivalenol, and deoxynivalenols, which are detrimental to livestock and humans.Contamination by these mycotoxins, in turn, imposes the risk of financial losses on growers. Furthermore, F. graminearuminfection presents a long-term agricultural concern as it overwinters in soil and infects corn planted in the next season. This combination of tissue damage, toxicity, and soil persistence makes F. graminearumone of the most destructive fungal pathogens to grain corn production.
While the threat of F. graminearumon grain corn production is well known, the interactions between F. graminearum and maize at the seedling stage are not well studied, and the influence of fungal and plant genotypes on disease severity remains elusive. Traditionally, GSB resistance level of a maize hybrid is indirectly measured as seedling standing rate in large scale field tests conducted by seed companies and agricultural research facilities. This method may not be optimal since seedling standing rate can be affected by many other soil-dwelling pathogens.Hence, in this project, we emphasize using diverse fungal isolates and commercial maize hybrids for evaluatingGSB severity of each combination of fungus-plant genotypes with a suite of quantitative traits under a controlled growth and inoculation condition. The agronomic values of these traits are then validated by calculating their correlations with seedling standing rate measured in the same maize seedling population inoculated withF. graminearum.
- Set up rigorous plant growth and artificial inoculation protocols in controlled growth chamber conditions;
- Evaluate the repeatability of various quantitative morphological, microscopic, biochemical, and molecular phenotypes induced by F. graminearum inoculation, as well as their variation among diverse maize and fungal genotypes;
- Calculate the statistical correlation between these fungus-induced phenotypic changes and lodging rate to establish a robust F. graminearum seedling blight resistance evaluation system applicable to diverse maize and fungal genotypes.
Plant growth and artificial inoculation protocol. We used four fungusisolates with contrasting pathogenicity and nine commercial maize hybrids kindly provided by Blue River Organic Seeds (Ames, IA) with variable maturation time and manufacturer-assessed seedling vigor. Maize seeds were surface-sterilized in 0.6% sodium hypochlorite for two minutes, rinsed in sterilized water, and soaked overnight. Seedswere then rolled in damped germination paper and kept upright for germination for 2-4 days untilthe primary roots wereapproximately 9 cm. Seedlings were immersed in one million spores/mL F. graminearum spore suspension harvested from a 7-day-old culture maintained on Potato Dextrose Agar. Fungus- and mock-inoculated seedlings were planted in calcinated clay (Turface MVP Athletics, Profile Product LLC., Buffalo Grove, IL). Fifteen technical replicates were planted for each combination of fungus isolate and maize hybrid.
Microscopic observation of F. graminearuminfected maize seedling roots. We grew a maize inbred line B73, and inoculated with F. graminearumstrain ZTE-2A, which constitutively expresses green fluorescence protein, using the protocol described above. Six days after inoculation, we are able to observe fungal hyphae growing superficially on seedling root surfaces with a fluorescence microscope.
Quantification of plant defense-related metabolites,mycotoxin, and defense-related gene expression. Benzoxazinoids are a class of indole-derived metabolites that are important for plant defense against various pathogens and insect herbivores. We extracted benzoxazinoids from mock- and fungus-inoculated maize seedling root six days after inoculation by soaking approximately 100 milligram of frozen and ground root tissue in 300 milliliter of 30% aqueous methanol with 0.1% formic acid extraction solvent for forty minutes. To estimate the concentration of major benzoxazinoid compounds in the extract, we spiked the extraction solvent with 25 nanomolar 2-benzoxalinone (Sigma Aldrich) as an internal standard. We analyzed the root extract through an Agilent 2695 High Performance Liquid Chromatography machine connected to a ultraviolet diode array detector (HPLC-DAD), and quantify the benzoxazinoid compounds by UV absorbance at 260 nanometer wavelength. Deoxynivalenol produced by F. graminearumcan be readily co-extracted with the 30% methanol solvent with benzoxazinoids. However, it cannot be detected or quantified from the same HPLC analysis, potentially because its concentration is lower than the detection limit. Therefore, we quantify the deoxynivalenol concentration in fungus-inoculated maize seedling root extract with a specific enzyme-linked immunosorbent assay (ELISA) kit (HelicaBiosystem) following manufacturer’s protocol.
To quantify growth and spread of F. graminearum, we measure expression of F. graminearum-specific genes by quantitative reverse transcriptase polymerase chain reactions (q-RT-PCR). We extract RNA from frozen and ground fungus-inoculated root tissue with Promega Wizard RNA Extraction kits (Promega), and quantified extracted RNA solution with a NanoDrop machine (Thermo Scientific). We then synthesize complementary DNA (cDNA) from 1 nanogram of RNA with MMLV Reverse Transcriptase kits (Clontech). We design q-RT-PCR primers targeting a maize actin gene and a F. graminearumalpha tubulin gene, and perform q-RT-PCR experiment with these primers and cDNA using SYBR Green reaction mix (Thermo Scientific). We run three technical replicates of each biological sample, and normalize expression of the F. graminearumalpha tubulin gene by that of the maize actin gene.
Morphological data collection and analysis. Fungus- and mock-inoculated seedlings were monitored for i) seedling height, ii) sheath length, and iii) emergence of the third leaf every two days for nine days after inoculation. In addition, nine days after inoculation, seedlings were removed from pots to measure their total root length usingan image analysis algorithm implemented in the RootReader 2D software. After measurement, seedlings were re-potted in the same Turface particles for measurement of shoot phenotypes at later time points. At 14 days post-inoculation, we counted the number of dead seedlings within eachfungus-inoculated population as a measurement of GSB severity. For seedling height, sheath length, and total root length, we calculated an average ratio of mock- and fungal-inoculated seedlings for each pair of maize hybrid-fungusisolate combination. To quantify the influence of F. graminearum inoculation on maize seedling development, we counted the proportion of seedlings with emerged third leaves among both mock- and fungus-inoculated seedlings, and calculated the difference between the two groups. We then calculate the pairwise linear regression relationship among these fungal-induced phenotypic changes, as well as lodging rate at fourteen days post-inoculation.
Evaluation of F. graminearumresistance based on fluorescence microscopy, biochemical markers, and gene expression are suboptimal. Fungal hyphae transformed with constitutive GFP expression are clearly visible at six days post-inoculation (Figure 1). However, quantification of fungal tissue based on fluorescence microscopy is laborious and limiting in throughput. This method became even more problematic when we started testing field-collected fungal isolates, which do not carry constitutive GFP marker, and have to be labeled by wheat germ agglutinin (WGA)-Alexa Fluor conjugates. This is because WGA is a non-specific chitin-binding molecule, which would add the Alexa Fluor marker to any fungal contaminant commonly found in our non-sterile growth conditions. While fluorescence microscopy method is primarily limiting in its quantification accuracy and throughput, quantification of biochemical and gene expression markers appear to beinconsistent in our pilot experiments. When repeating metabolite concentration and gene expression measurement across multiple independent experiments under supposedly identical experimental conditions. In addition to these significant flaws, we also recognize that fluorescence microscopy, HPLC, and q-RT-PCR are often inaccessible to farmers, and pre-emptive screening based on these methods may not be economically viable. Therefore, we decide to look for more accessible traits that can be easily quantified with high accuracy and repeatability.
Fusarium graminearum infestation leads to quantifiable morphological changes in maize seedlings both at individual and population level. Across the diverse maize hybrid lines and F. graminearum isolates tested, we found that F. graminearum inoculation can lead to accelerated seedling development as evidenced by elongation of shoot and sheath of individual seedlings and earlier emergence of third leaves in F. graminearum-inoculated populations (Figure 2). In contrast, F. graminearum inoculation leads to severe reduction of total root length, even though root lesion, browning and shrinkage remains local (Figure 3). Together, these systemic changes in plant morphology suggest widespread plant defense responses upon F. graminearum inoculation and provide potential target phenotypes to use as quantitative predictors of GSB severity.
Gibberella seedling blight severity is primarily determined by F. graminearum but not maize genotype. We measured GSB severity as percent lodging in each maize hybrid population fourteen days after inoculation with different F. graminearum isolates. Analysis of variance (ANOVA) of these lodging rate data demonstrates that the genotype of the F. graminearum isolates has a significant impact on the Gibberella seedling blight severity, whereas the genotype of maize hybrids is not a significant source of variance, nor are the maturity time or the manufacturer assessed seedling vigor (Table 1). Consistent with this analysis, we find Gz835 to be a highly pathogenic fungal isolate, leading to 78% lodging across all maize hybrids tested. On the other hand, Gz941 appears to be only moderately pathogenic (20% lodging), and Gz014 and Gz829 are between the extremes (Figure 4A). Average lodging rate of each maize hybrid ranges from 35% to 65%, but is significantly variable within each hybrid, consistent with the ANOVA result (Figure 4B).
These data are consistent with existing consensus that current maize commercial hybrids do not have any genetic source of F. graminearum resistance, and the severity GSB would be highly variable, depending on the pathogenicity of the dominant fungal isolate of a particular year. This finding further underlines the necessity of developing an accurate evaluation and prediction system for GSB severity.
Fusarium graminearum induced plant morphological changes are quantitatively correlated with Gibberella seedling blight severity as well as with each other. To investigate the predictive power of F. graminearum-induced morphological changes for GSB severity, we perform linear regression analysis of the extent of these changes at different time point and two-week seedling lodging rate. Phenotypic changes in seedling height at the later time point (9 days post-inoculation, DPI) are strongly correlated with seedling lodging rate measured five days later (Figure 5A). Similarly, third leaf emergence at 7 DPI is also significantly correlated with the seedling lodging rate (Figure 5B). The same phenotype cannot be quantified at 9 DPI because both mock- and fungus-inoculated seedlings have fully emerged third leaves at this time point. Surprisingly, the extent of seedling height elongation as early as 3 DPI is significantly correlated with the seedling lodging rate measured at 14 DPI (Figure 5C). All of the significant correlations between shoot morphological changes and lodging rate are positive, suggesting that more severe F. graminearum-induced morphological changes are predictive of more severe GSB later on. Furthermore, we observed that F. graminearum-induced root reduction is significantly correlated with morphological changes in the shoot, but is not correlated with seedling lodging rate (Table 2). To our surprise, F.graminearum-induced shoot elongation and root reduction are negatively correlated with each other, suggesting these two fungal-induced morphological changes may be regulated by distinct physiological mechanisms.
Since we have identified genotype of F. graminearum as a major influencer of GSB severity, we also investigated the effect of F. graminearum genotype on the correlation between fungus-induced morphological changes and later GSB severity. The correlations generally hold true, though they are less robust, with the exception of the highly pathogenic Gz835, where the linear relationships between fungus-induced morphological changes and lodging rate are abolished by very high overall lodging rate (Figure 5).
From these results, we conclude that, under our controlled growth conditions, F. graminearum-induced acceleration in shoot growth and development is an accurate predictor of GSB severity later in development, such that abnormally fast shoot growth can suggest higher seedling lodging rate from F. graminearum infection. F. graminearum inoculation is also observed to induce root reduction. Though this morphological change is not correlated with later seedling lodging rate, we expect that severe reduction in total root length and hence surface area to significantly impair the seedlings’ ability to acquire water and nutrients from soil, and negatively impact plant health.
From this project, we have found that F. graminearum infection in maize seedlings can lead to accelerated shoot development and reduced root growth prior to seedling lodging. These morphological changes are similar to the bakanae disease found in rice seedlings infected with Fusarium verticilioides, a related fungal pathogen. We further demonstrated the quantitative correlation between the extent of abnormal shoot development at an early infection stage and later GSB severity measured by seedling lodging rate. This relationship, if it holds true under field conditions, could provide an easily accessible method for farmers to forecast GSB severity. The value of this forecasting tool is further underlined by our confirmatory finding that current commercial maize hybrids do not have a genetic source of resistance against F. graminearum, and the GSB severity is largely variable from year to year depending on the pathogenicity of fungal isolates and environmental permissiveness. By adopting this forecasting tool, farmers could have a more reasonable expectation of GSB severity in a specific growing season, reduce expensive and environment-unfriendly fungicide application when unnecessary, and achieve more sustainable farming practices.
In addition to the direct practical value, our findings also provide an important and novel phenotyping tool for researchers studying F. graminearum resistance in maize. The F. graminearum-induced morphological changes we find can serve as an easily quantifiable proxy for measuring F. graminearum resistance in different maize genotypes, which is essential for reliable genetic mapping studies. As an example, I have used F. graminearum-induced shoot elongation and root reduction to screen a collection of maize inbred lines and found one of them to be highly insensitive to F. graminearum inoculation when compared to most other inbreds. This result has laid the foundation for follow-up comparative metabolomics and metabolite trait mapping studies.
Education & Outreach Activities and Participation Summary
We plan to include data and results generated from this project in my PhD thesis, and publish them as either an integrated component of a more comprehensive study of maize-F. graminearum interactions, or an independent agronomic study. My undergraduate lab assistants, Justin Bae and Jessica Aduwo, and I have presented these data as both posters and oral presentations in various occasions, including Cornell and Boyce Thompson Institute Summer Research Intern Presentations, in Plant Biology Graduate Field talks, and the Grass Research Group meeting at Cornell University. These data also will be presented as a poster at the upcoming Maize Genetics Conference in March 2016.
We took the opportunity of this project to get undergraduate students extensively involved in the experimental design, data collection and analysis, and result interpretation process. Through these hands-on laboratory experiences and interactions with plant scientists, we hope our students have developed the appreciation of the significance of research devoted to sustainable agriculture, as well as plant research processes.
Since we finished the majority of our data collection from growth chamber experiments towards the end of summer 2015, we did not have the opportunity to communicate our results with the broader farming community as the GSB onset time in the field has already passed. For the coming growing season of 2016, we are proactively seeking opportunities to communicate our findings with farmers through the extensive agricultural outreach network of Cornell University in forms of printouts or in-field information sessions.
As the research portion of this project has just been completed and farmer adoption has not yet been assessed, it is not possible to conduct an economic analysis.
As the research portion of this project has just been completed, it is not yet possible to assess farmer adoption.
Areas needing additional study
The major perspective that requires further study is the influence of environmental conditions on the relationship between F. graminearum-induced morphological changes and GSB disease severity. To better focus on the effects of F. graminearum infection, we purposefully chose to perform all our experiments under controlled growth conditions with artificial inoculation and acontrolled fungal load. We recognize that underfield environments plant growth is influenced by a wide array of abiotic and biotic stress factors, and it remains to be tested whetheri) the impact of F. graminearum infection can be clearly identified, and ii) the relationship between F. graminearum-induced morphological changes and GSB severity would still hold true in such a highly variable environment. In our original project proposal, we proposed to address this problem by sampling seedling growth rate in various field locations across NY and testing for correlation with later reported GSB severity (Objective 3). However, we have not been able to collect data for this analysis up to this date because we didn’t finish the laboratory phase of the project before the only field season (Summer 2015) within the span of grant period. Since we now understand that seedling height is the particular phenotype to look for, we are hoping to collect these data in the upcoming growing season of 2016.