Genome-wide prediction of activating regulatory elements in rice by combining STARR-seq with FACS

Self-transcribing active regulatory region sequencing (STARR-seq) is widely used to identify enhancers at the whole-genome level. However, whether STARR-seq works as efficiently in plants as in animal systems remains unclear. Here, we determined that the traditional STARR-seq method can be directly applied to rice (Oryza sativa) protoplasts to identify enhancers, though with limited efficiency. Intriguingly, we identified not only enhancers but also constitutive promoters with this technique. To increase the performance of STARR-seq in plants, we optimized two procedures. We coupled fluorescence activating cell sorting (FACS) with STARR-seq to alleviate the effect of background noise, and we minimized PCR cycles and retained duplicates during prediction, which significantly increased the positive rate for activating regulatory elements (AREs). Using this method, we determined that AREs are associated with AT-rich regions and are enriched for a motif that the AP2/ERF family can recognize. Based on GC content preferences, AREs are clustered into two groups corresponding to promoters and enhancers. Either AT- or GC-rich regions within AREs could boost transcription. Additionally, disruption of AREs resulted in abnormal expression of both proximal and distal genes, which suggests that STARR-seq-revealed elements function as enhancers in vivo. In summary, our work provides a promising method to identify AREs in plants.

Automated summary

In this study, the researchers found that the traditional STARR-seq method can be applied to rice protoplasts to identify enhancers, although with limited efficiency. They optimized the STARR-seq method by using fluorescence activating cell sorting (FACS) and minimizing PCR cycles and duplicates. The study revealed that AREs are associated with AT-rich regions and contain a motif recognized by the AP2/ERF family. AREs can be clustered into two groups corresponding to promoters and enhancers based on GC content preferences. Disruption of AREs resulted in abnormal gene expression. This work provides a promising method to identify AREs in plants.