The CRF domain defines cytokinin response factor proteins in plants.

backgroundCytokinin Response Factors (CRFs) are a small subset of AP2/ERF transcription factor genes shown in Arabidopsis to regulate leaf development as part of the cytokinin signal transduction pathway. This study examines the phylogenetic distribution of CRF genes in other plant species, and attempts to identify the extent of sequence conservation and potential gene function among all CRF genes.

resultsWe identified CRF genes in representatives of all major land plant lineages, including numerous flowering plant taxa in addition to the model systems in which ERF genes have been catalogued. Comparative analysis across this broader sampling has identified strongly conserved amino acid motifs other than the AP2/ERF domain for all CRF proteins as well as signature sequences unique to specific clades of CRF genes. One of these motifs, here designated as the CRF domain, is conserved in and unique to CRF proteins distinguishing them from related genes. We show that this novel domain of approximately 65 amino acids is found in CRF proteins from all groups of land plants and only in CRF genes. Phylogenetic analyses suggest that the evolution of CRF genes has included numerous duplication events. In this phylogenetic context we examine protein evolution including the gain and loss of accessory domains, correlate these molecular evolutionary events with experimental data on cytokinin regulation and speculate on the function and evolution of the CRF domain within AP2/ERF transcription factor proteins. We also tested a prediction drawn from the phylogenetic analyses that four CRF domain containing genes from Tomato, previously unexamined for cytokinin response, are transcriptionally inducible by cytokinin, supporting the link between CRF genes, CRF-specific domains and cytokinin regulation.

conclusionCRF genes can be identified in all lineages of land plants, as a distinct subset of AP2/ERF proteins containing a specific and unique CRF domain. The CRF domain can be used to identify previously unclassified predicted genes or genes identified only as members of the AP2/ERF protein family. CRF domain presence and phylogenetic relatedness to known Arabidopsis CRF genes predicts gene function to some extent.