Identifying genes that regulate natural variation for maize leaf cuticular wax abundance and cuticular evaporation rate
Maize is a critical component of animal feed, biofuel production, and, most importantly, our food supply. Yet, maize yield is significantly threatened by drought and thus the need for drought tolerant maize lines is substantial. Maize leaf cuticular evaporation (CE) rate, the water evaporation through the cuticle and incompletely sealed stomata, is a major source of water loss during nights and periods of drought. Cuticular wax, which belongs to the cuticle, or the hydrophobic film covering the leaf epidermis, has been hypothesized to be associated with CE rate. Therefore, identifying genes associated with CE rate and cuticular waxes, as well as understanding how they relate, is critical to improving maize drought tolerance. To study this, CE rate was evaluated for 320 maize inbred lines in the Wisconsin Diversity Panel planted in San Diego, California in 2018. Cuticular waxes were extracted from the 60 maize lines with the most extreme high or low CE rates, and wax samples were analyzed for 59 wax components. 3’ RNA sequencing was additionally conducted using the base section of developing adult leaves. The wax data was then normalized using a box-cox transformation after outlier removal. Gene expression patterns due to experimental design and other hidden factors were extracted from the gene expression data using probabilistic estimation of expression residuals (PEER), and PEER residuals of gene expression data were used for transcriptome-wide association studies (TWAS). TWAS were performed using linear models to test associations between each wax component and gene expression levels, with wax population structures adjusted by principal components. The number of principal components used in each model was determined by Bayesian information criterion (BIC). The resulting TWAS indicated eight genes significantly associated with six wax components at a false discovery rate (FDR) of 5%, with one promising candidate gene encoding for a cycloartenol synthase, an enzyme with suggested involvement in alicyclic synthesis. Furthermore, to estimate the extent of correlation between wax components and CE rate, cuticular wax abundancy was used to predict CE rate by implementing a random forest generated in R. The average predictive ability, estimated by the Pearson’s correlation coefficient between the observed CE rate and the predicted CE rate, was 0.32 (r = 0.12 to 0.47), thus demonstrating that variation in wax composition influences CE rate. Additionally, a genome-wide association study (GWAS) has been conducted on CE rate. To further validate the biological function of a candidate gene identified in GWAS, we planted three UniformMu maize lines with insertions disruptive of this gene, as well as a background wild type line in Aurora, NY. Leaf tissue was collected, and DNA was subsequently extracted. Genotyping of samples revealed that no samples were homozygous mutants, yet heterozygous mutants were identified. Heterozygous mutants will thus be self-pollinated to produce homozygous mutant offspring, which will later be used for CE rate and wax composition evaluation. Genes identified in this study can provide valuable information towards understanding leaf cuticle development and improving drought tolerance in maize.
I’ve had a fantastic experience as a BTI intern in the Gore lab this summer. Most notably, I learned what I wasn’t expecting to learn— bioinformatics. I had started out intimidated by coding and with only basic capabilities in R, but I am now so grateful to have become friends with this software. I have grown a real appreciation for this impressive tool. It has been a fulfilling challenge that taught me to be persistent and think creatively, and I have learned so many statistical methods that were completely foreign to me prior. I also now have a greater understanding of what biological research entails; while I had hardly heard of bioinformatics before, it is profoundly clear to me now just how critical it is. I’m very thankful to have developed this crucial skill, with many thanks to my incredibly patient and helpful mentor Meng Lin (and not to mention plentiful R help forums!). Additionally, I’m so appreciative to now feel more independent and comfortable preforming laboratory techniques, and to have seen what a life in research is like. I will miss working here, the people who I’ve met, and living in Ithaca, but I am excited to take what I have learned here and apply it to future research I hope to involve myself in. And just as I’d hoped for upon arriving here, I’m glad to be able to take away all this insider-like knowledge of the wonderful little details of maize!