Kakei Yusuke, Masuda Hiroshi, Nishizawa Naoko K, Hattori Hiroyuki, Aung May Sann
Institute of Vegetable and Floriculture Science, Research Center for Agricultural Information Technology, National Agriculture and Food Research Organization, Ibaraki, Japan.
Faculty of Bioresource Sciences, Department of Biological Production, Akita Prefectural University, Akita, Japan.
Front Plant Sci. 2021 Jun 2;12:660303. doi: 10.3389/fpls.2021.660303. eCollection 2021.
Iron (Fe) excess is a major constraint on crop production in flooded acidic soils, particularly in rice cultivation. Under Fe excess, plants activate a complex mechanism and network regulating Fe exclusion by roots and isolation in various tissues. In rice, the transcription factors and -regulatory elements (CREs) that regulate Fe excess response mechanisms remain largely elusive. We previously reported comprehensive microarray analyses of several rice tissues in response to various levels of Fe excess stress. In this study, we further explored novel CREs and promoter structures in rice using bioinformatics approaches with this microarray data. We first performed network analyses to predict Fe excess-related CREs through the categorization of the gene expression patterns of Fe excess-responsive transcriptional regulons, and found four major expression clusters: Fe storage type, Fe chelator type, Fe uptake type, and WRKY and other co-expression type. Next, we explored CREs within these four clusters of gene expression types using a machine-learning method called microarray-associated motif analyzer (MAMA), which we previously established. Through a comprehensive bioinformatics approach, we identified a total of 560 CRE candidates extracted by MAMA analyses and 42 important conserved sequences of CREs directly related to the Fe excess response in various rice tissues. We explored several novel -elements as candidate Fe excess CREs including GCWGCWGC, CGACACGC, and Myb binding-like motifs. Based on the presence or absence of candidate CREs using MAMA and known PLACE CREs, we found that the Boruta-XGBoost model explained expression patterns with high accuracy of about 83%. Enriched sequences of both novel MAMA CREs and known PLACE CREs led to high accuracy expression patterns. We also found new roles of known CREs in the Fe excess response, including the DCEp2 motif, IDEF1-, Zinc Finger-, WRKY-, Myb-, AP2/ERF-, MADS- box-, bZIP and bHLH- binding sequence-containing motifs among Fe excess-responsive genes. In addition, we built a molecular model and promoter structures regulating Fe excess-responsive genes based on new finding CREs. Together, our findings about Fe excess-related CREs and conserved sequences will provide a comprehensive resource for discovery of genes and transcription factors involved in Fe excess-responsive pathways, clarification of the Fe excess response mechanism in rice, and future application of the promoter sequences to produce genotypes tolerant of Fe excess.
铁（Fe）过量是淹水酸性土壤中作物生产的主要限制因素，尤其是在水稻种植中。在铁过量的情况下，植物会激活一种复杂的机制和网络，调节根系对铁的排斥以及铁在不同组织中的隔离。在水稻中，调节铁过量响应机制的转录因子和顺式作用元件（CREs）在很大程度上仍不清楚。我们之前报道了对几种水稻组织在不同水平铁过量胁迫下的全面微阵列分析。在本研究中，我们利用这些微阵列数据，通过生物信息学方法进一步探索了水稻中的新型CREs和启动子结构。我们首先进行网络分析，通过对铁过量响应转录调控因子的基因表达模式进行分类来预测与铁过量相关的CREs，发现了四个主要的表达簇：铁储存类型、铁螯合剂类型、铁吸收类型以及WRKY和其他共表达类型。接下来，我们使用一种名为微阵列相关基序分析器（MAMA）的机器学习方法，该方法是我们之前建立的，来探索这四个基因表达类型簇中的CREs。通过综合生物信息学方法，我们共鉴定出560个通过MAMA分析提取的CRE候选序列，以及42个与水稻不同组织中铁过量响应直接相关的CRE重要保守序列。我们探索了几个新型元件作为候选铁过量CREs，包括GCWGCWGC、CGACACGC和类Myb结合基序。基于使用MAMA和已知PLACE CREs的候选CREs的存在与否，我们发现Boruta-XGBoost模型以约83%的高精度解释了表达模式。新型MAMA CREs和已知PLACE CREs的富集序列都导致了高精度的表达模式。我们还发现了已知CREs在铁过量响应中的新作用，包括铁过量响应基因中含DCEp2基序、IDEF1-、锌指、WRKY、Myb、AP2/ERF、MADS盒、bZIP和bHLH结合序列的基序。此外，我们基于新发现的CREs构建了一个调节铁过量响应基因的分子模型和启动子结构。总之，我们关于铁过量相关CREs和保守序列的研究结果将为发现参与铁过量响应途径的基因和转录因子、阐明水稻中铁过量响应机制以及未来应用启动子序列培育耐铁过量基因型提供全面的资源。
