Parallel inactivation of the APETALA3-3 genes contributed to the independent petal losses within the buttercup family (Ranunculaceae)
Rui Zhang^1,2,, Chunce Guo^1,2,, Wengen Zhang^1,2,, Peipei Wang^1,2, Lin Li^1,2, Xiaoshan Duan^1,2,3, Qinggao Du³, Liang Zhao³, Hongyan Shan³, Scott A. Hodges⁴, Elena M. Kramer⁵, Yi Ren^3,, Hongzhi Kong¹

^aState Key Laboratory of Systematic and Evolutionary Botany, Institute of Botany, the Chinese Academy of Sciences, Beijing, 100093, China;
^bGraduate University of Chinese Academy of Sciences, Beijing, 100049, China;
^cCollege of Life Sciences, Shaanxi Normal University, Xi’an, 710062, China.

Abstract
Petals show enormous diversity in appearance and function, and play key roles in angiosperm evolution. Petals have been lost multiple times, giving rise to the occurrences of apetalous flowers in different lineages. The underlying mechanisms for petal losses are still unclear, yet recent studies suggested that petal losses within the buttercup family (Ranunculaceae) are strongly correlated with silencing or down-regulation of a single gene, APETALA3-3 (AP3-3). To understand the mechanisms behind this, we investigated the expression patterns and sequence features of the AP3-3 orthologs from eleven representative taxa. We found that the AP3-3 orthologs of petalous plants are usually specifically and highly expressed in petals while those of apetalous ones are either not expressed or expressed at very low levels. In Nigella, insertion of a transposable element into the second intron has led to the inactivation of the gene and the generation of the apetalous double-flower mutant. In Thalictrum, the gene is no longer existing, suggestive of gene loss. In Beesia, the gene is present but defective: two deletions in the last exon have led to the occurrence of a pre-mature stop codon and down-regulation of the gene. In Enemion, a short deletion in the otherwise highly conserved promoter region has likely disrupted the expression process. In Clematis, the gene does not show obvious defects in coding or regulatory regions but has evolved neutrally. Taken together, these results suggest that AP3-3 is a key regulator of petal development but has been silenced or down-regulated in different lineages by different mechanisms.