Extreme restructuring of cis-regulatory regions controlling a deeply conserved plant stem cell regulator

Abstract

A striking paradox is that genes with conserved protein sequence, function and expression pattern over deep time often exhibit extremely divergent cis-regulatory sequences. It remains unclear how such drastic cis-regulatory evolution across species allows preservation of gene function, and to what extent these differences influence how cis-regulatory variation arising within species impacts phenotypic change. Here, we investigated these questions using a plant stem cell regulator conserved in expression pattern and function over [~]125 million years. Using in-vivo genome editing in two distantly related models, Arabidopsis thaliana (Arabidopsis) and Solanum lycopersicum (tomato), we generated over 70 deletion alleles in the upstream and downstream regions of the stem cell repressor gene CLAVATA3 (CLV3) and compared their individual and combined effects on a shared phenotype, the number of carpels that make fruits. We found that sequences upstream of tomato CLV3 are highly sensitive to even small perturbations compared to its downstream region. In contrast, Arabidopsis CLV3 function is tolerant to severe disruptions both upstream and downstream of the coding sequence. Combining upstream and downstream deletions also revealed a different regulatory outcome. Whereas phenotypic enhancement from adding downstream mutations was predominantly weak and additive in tomato, mutating both regions of Arabidopsis CLV3 caused substantial and synergistic effects, demonstrating distinct distribution and redundancy of functional cis-regulatory sequences. Our results demonstrate remarkable malleability in cis-regulatory structural organization of a deeply conserved plant stem cell regulator and suggest that major reconfiguration of cis-regulatory sequence space is a common yet cryptic evolutionary force altering genotype-to-phenotype relationships from regulatory variation in conserved genes. Finally, our findings underscore the need for lineage-specific dissection of the spatial architecture of cis-regulation to effectively engineer trait variation from conserved productivity genes in crops. Author summaryWe investigated the evolution of cis-regulatory elements (CREs) and their interactions in the regulation of a plant stem cell regulator gene, CLAVATA3 (CLV3), in Arabidopsis and tomato. Despite diverging [~]125 million years ago, the function and expression of CLV3 is conserved in these species; however, cis-regulatory sequences upstream and downstream have drastically diverged, preventing identification of conserved non-coding sequences between them. We used CRISPR-Cas9 to engineer dozens of mutations within the cis-regulatory regions of Arabidopsis and tomato CLV3. In tomato, our results show that tomato CLV3 function primarily relies on interactions among CREs in the 5 non-coding region, unlike Arabidopsis CLV3, which depends on a more balanced distribution of functional CREs between the 5 and 3 regions. Therefore, despite a high degree of functional conservation, our study demonstrates divergent regulatory strategies between two distantly related CLV3 orthologs, with substantial alterations in regulatory sequences, their spatial arrangement, and their relative effects on CLV3 regulation. These results suggest that regulatory regions are not only extremely robust to mutagenesis, but also that the sequences underlying this robustness can be lineage-specific for conserved genes, due to the complex and often redundant interactions among CREs that ensure proper gene function amidst large-scale sequence turnover.