Evaluation and prediction of Fusarium graminearum resistance in New York grain corn hybrids

Evaluation and prediction of Fusarium graminearum resistance in New York grain corn hybrids

Summary

Starting August 2014, I devoted the first four months of this project to three major aspects tightly related to the key features of this project. In my proposal, I highlighted the use of artificial inoculation at seedling stage in controlled growth chamber condition for more uniform fungal induction. To acheive this goal, I tested five different inoculation and plant growth methods for optimal outcome. This optimization process is essential for consistent phenotypic observations, and hence the overall value of this project. A second important feature of this project is the inclusion of diverse maize lines and fungal isolates. To gather the most relevant plant materials, I set up a connection with an organic maize seed company to receive the most up-to-date germplasm (as oppose to the older hybrids listed in the original proposal). Field-collected F. graminearum isolates were supplied by Dr. Bergstrom’s laboratory as proposed. Finally, I emphasized the importance of using multiple phenotypes in characterizing the pathogenic interaction between maize seedlings and F. graminearum, and listed many potentially useful phenotypes using microscopic, biochemical or molecular observations. The feasibility and repeatability of these phenotypes, however, were not fully evaluated at the time of proposal, and hence require further testing.

Objectives/Performance Targets

1. Establish effective plant growth and fungal inoculation protocols;

2. Gather required maize lines and fungal isolates;

3. Pilot tests for natural variation and repeatability of proposed phenotypes.

Accomplishments/Milestones

1. Plant growth and fungal inoculation method

I tested different plant growth and fungal inoculation methods using various potting medium (e.g. hydroponic, sand, and gravels) and fungal inoculum (e.g. spore suspension and agar blocks) for uniform fungal infection. At this point, the most successful method is directly immersing roots of sand-grown seedlings in F. graminearum spore suspension for 1 hour before re-planting the seedlings back into the same potting medium.

2. Gather required plant and fungal materials

Information of organic maize hybrids and field-collected fungal isolates are summarized in Table 1 & 2 in the supplemental file attached.

3. Pilot tests for natural variation and repeatability of proposed phenotypes.

My project uses multiple phenotypes to characterize the pathogenic interaction between maize seedlings and F. graminearum. It is hence important to explore if these phenotypes can be consistently observed in my experimental setup and are variable among the maize genotypes I’m proposing to screen. To this end, I grew eight maize inbred lines (n>4 per line) in germination pouches, drip inoculated seedling roots with  F. graminearum spore suspension with 0.02% phytoagar seven days post-planting, and harvested leaves and roots separately three days after inoculation. For this preliminary experiment, I used a transformed F. graminearum strain ZTE-2A containing a jellyfish green fluorescence protein (GFP) This transformed F. graminearum strain ZTE-2A constitutively expresses GFP, and hence allows visual confirmation of infection status using epifluorescence microscopy. Total benzoxazinoids are extracted from F. graminearum ZTE-2A and mock- (0.02% phytoagar only) inoculated maize seedling leaves and roots, partitioned with High Performance Liquid Chromatography (HPLC), and quantified based on ultraviolet (UV) light absorption using a photodiode array detector.

Results from this preliminary experiment revealed significant natural variation in DIMBOA-glucoside and HDMBOA-glucoside among the eight NAM founder lines in an organ-specific manner (Figure 1). In roots, DIMBOA-glucoside level is unaffected by F. graminearum inoculation in six of eight lines tested (including the common NAM parent B73), but significantly reduced in CML322 and accumulated in Ki11 (Figure 1A). HDMBOA-glucoside level in roots, however, significantly decreases in six of the eight lines tested, while remains unchanged in Oh7b and insignificantly reduced in B73 (Figure 1B). In leaves, only five of the eight lines tested have sufficient replicates for meaningful statistical comparisons. Among these five lines, DIMBOA-glucoside level is only accumulated in F. graminearum inoculated CML228 (Figure 1C), while significant F. graminearum-induced reduction of HDMBOA-glucoside is only observed in CML322 (Figure 1D). Juxtaposing induced changes of DIMBOA-glucoside and HDMBOA-glucoside in roots and leaves of the same maize genotype further underlines the organ-specific nature of benzoxazinoids dynamics in response to F. graminearum inoculation. In summary, these preliminary results strongly support that F. graminearum-induced changes in different benzoxazinoids are significantly variable across selected NAM founder lines, and this variability should be conserved and further expanded across the full maize diversity panel I’m proposing to study (Table 1; see Approach and Methods). Hence, this natural variation can be exploited to achieve my proposed objectives.

In the same experiment, an unexpected peak was observed and only observed in CML322 leaf tissue harvested from F. graminearum inoculated seedlings (Figure 2). Examination of the UV absorbance profile of this peak suggests that it is possibly a mixture of three or more compounds, and it is unclear if one or more of them are necessarily related to benzoxazinoids. It will be very interesting to further partition this mixture of compounds, identify them with mass spectrometry, attempt to associate them with particular functions as fungal-induced metabolites, and adopt them as additional phenotypes in the genetic mapping portion of this project.

I also searched for absorbance peaks possibly corresponding to DONs that are produced by F. graminearum strain PH-1 (Trail and Common, 2000), the wild type progenitor of ZTE-2A used in this preliminary study. However, no peak has been consistently observed with a presence/absence difference between fungal- and mock-inoculated root samples, which is the expected pattern of any fungal-secreted metabolite. This failure of detection may be an artifact of incompatible extraction and analytic method, since previous literature suggests that gas chromatography-mass spectrometry (GC-MS) and specific enzyme-linked immunosorbent assays (ELISA) are more reliable quantification methods for DONs, whereas HPLC-based methods usually require prior immunoprecipitation cleanup.Meanwhile, q-RT-PCRanalysis of B73 root samples with confirmed fungal α-tubulin presence shows active FgTRI5 expression (Figure 3). This ambiguous result hence presses the need of another dedicated assay for clearer DON detection and quantification.

For microscopic observations, however, results were highly inconsistent among replicates of the same genotype. This may be a result of sub-optimal inoculation method. Furthermore, microscopic phenotyping proved to be a very laborious process, which is not desirable for a project focusing on the breadth of natural variation. Taken together, I decided to drop out the microscopic components of this project and better focus on biochemical and molecular phenotypes.

• ZhouGNE14-092AR

Impacts and Contributions/Outcomes

As this project is still at very early stage, further experimentation and analysis are required before any impact or outcome can become obvious.

Collaborators: