Screening and Application of Biomarkers Related to Maize Nitrogen Status

The transcriptional levels of selected genes, referred as biomarkers, have been widely applied in clinical diagnosis processes. They were yet rarely used in agricultural cultivation for determining the nutrient statues in maize.

This study aims to explore the genes that can be used as biomarkers to reflect the nitrogen abundance in maize, so as to help the precise application of nitrogen fertilizer.

Based on the data of gene chip and RNA-Seq under different nitrogen treatments, we chose the genes with high transcriptional abundance in response to N fluctuation as candidates. These genes were further screened by qRT-PCR and Kjeldahl methods using maize materials with different genotypes under different nitrogen treatments. The generalized linear models for predicting nitrogen status were constructed to accurately indicate the nitrogen nutrition status of maize.

Firstly, we selected ten candidate genes with high expression level that are responsible for N fluctuation. Secondly, we found eight candidate genes that are differentially expressed under N treatment; Next, twenty-seven inbred and four hybrid lines covering a rich array of genetic diversity were selected to screen the candidate genes, and found that four genes stably expressed in different genotypes of maize. The expression abundance difference of these four genes were significantly correlated with total nitrogen content in panicle leaves through correlation analysis (R² was greater than 0.6) with sufficient nitrogen and limited nitrogen treatment in thirty materials. According to the above results, these four genes can be used as nitrogen response biomarkers to indicate maize nitrogen status. The two-genes, three-genes and four-genes models were constructed by these four biomarker genes for predicting nitrogen status. The three-genes model was composed of Zm000011d024281 (X₂), Zm000011d039049 (X₃) and Zm000011d037680 (X₄) were the most useful model for predicting the nitrogen status of maize plants, and the functional relationship was Y=1.143+0.017X₂-0.302X₃+0.017X_4. Finally, the prediction function of the three-genes model was verified with six hybrids planted in the field. The results show that the three-genes model can accurately diagnose the nitrogen nutrition status of maize planted in the field environment.

We explored and verified four biomarker genes highly responsive to maize nitrogen status. The three-genes model works best in predicting the maize nitrogen nutrition status. The development of the biomarker can effectively and real-timely monitor the nitrogen status of maize plants, thus is helpful for optimizing the use of nitrogen fertilizer, thereby maximize the crop yield at the lowest cost.