Next Generation Sequencing of Sorted pWER::GFP Marker Line Enables Quantitative Analysis of Arabidopsis Epidermal Root Transcriptomes in Iron Deficient Conditions

Purpose: Identify novel genes and relationships of a temporal transcriptional landscape in Arabidopsis epidermal root cells under iron deficiency stress Results: We identified 1,709 novel iron responsive genes which would have been missed had we identified DEGs with respect to the 0hr time point using the whole root or epidermis data. Conclusion: Gene set comparisons allowed us to identify organ-level response versus cell-specific and to parse out developmental from iron deficiency-induced response. We identified several new iron deficiency-induced and repressed TFs and implemented a novel sampling method to generate an epidermis specific GRN that incorporates gene-level expression while maintaining computational efficiency. Using cis-elements in the promoters of DEGs, we were also able to find plausible regulators that would not have been found by exclusively looking at differentially expressed genes. The new transcriptional data provides clarity to our current understanding of root responses to -Fe triggered by the first layer of “defense,” the root epidermis. Overall design: RNA-Seq was performed on protoplast collected from GFP sorted epidermis marker line (pWER::GFP; WEREWOLF; AT5G14750; Lee & Schiefelbein, 1999). Arabidopsis roots were collected on the 7th day of growth on +Fe media, equating to 0hr. After 7days, seedlings were transfer to +Fe and -Fe media, an additional 6hr, 12hr, 18hr, 24hr, 30hr, and 36hr of growth occurred. After which, roots were harvested from both +Fe and -Fe conditions for protoplast collection, followed by fluorescence-activated cell sorting in triplicates. RNA was extracted from FACS epidermal protoplast using RNAeasy Plant RNA Purification Kit (Qiagen). cDNA synthesis and amplification were performed using the SMARTer Low Input RNA Kit and sequenced using an Illumina HiSeq 2500 sequencing machine, with 125bp single end reads. Adapters and low quality reads were filtered out using fastQC and fastq-mcf. Clean reads were mapped against the TAIR 10 reference genome using the tophat2. RPKM were acquired using RSubread, followed by edgeR to identify differentially expressed genes (DEGs), using a maximum false discovery rate (FDR) of 0.05 and minimum log fold change threshold of 0.75.