Widespread transcription control from within transcribed regions in plants

We synthesized 12,000 sequences, each 160 bp long, derived from either 40-200 bp upstream or 40-360 bp downstream of the TSS of highly expressed genes in Arabidopsis, excluding the TSS region itself. Downstream-derived sequences included both exons and introns, but we removed donor and acceptor splicing sites from them. We inserted these sequences on either side of the TSS of a reporter GFP. We also added constructs with no insertions as controls. The insertion site downstream was within an intron to avoid altering the mature mRNA, thus minimizing the effect on mRNA stability. For quantification, we generated multiple copies of each construct, each containing a 15 bp random barcode within the transcript. We read these via next-generation sequencing, without using fluorescence.We employed two different backbone structures to provide varied genetic contexts. The first used the 35S minimal promoter, popular for ectopic expression in plants. The second involved a 700 bp promoter and 5' UTR from the Arabidopsis TRP1 gene, previously employed for studying intron effects on gene expression. To derive conclusions relevant to a broad range of flowering plants, we quantified the synthetic libraries in four species: Arabidopsis, tomato, maize (using leaf protoplasts), and N. benthamiana (through leaf infiltration).