Functional Divergence of APETALA1 and FRUITFULL is due to Changes in both Regulation and Coding Sequence

Abstract

Gene duplications are prevalent in plants, and functional divergence subsequent to duplication may be linked with the occurrence of novel phenotypes in plant evolution. Here, we examine the functional divergence of Arabidopsis thaliana APETALA1 (AP1) and FRUITFULL (FUL), which arose via a duplication correlated with the origin of the core eudicots. Both AP1 and FUL play a role in floral meristem identity, but AP1 is required for the formation of sepals and petals whereas FUL is involved in cauline leaf and fruit development. AP1 and FUL are expressed in mutually exclusive domains but also differ in sequence, with unique conserved motifs in the C-terminal domains of the proteins that suggest functional differentiation. To determine whether the functional divergence of AP1 and FUL is due to changes in regulation or changes in coding sequence, we performed promoter swap experiments, in which FUL was expressed in the AP1 domain in the ap1 mutant and vice versa. Our results show that FUL can partially substitute for AP1, and AP1 can partially substitute for FUL; thus, the functional divergence between AP1 and FUL is due to changes in both regulation and coding sequence. We also mutated AP1 and FUL conserved motifs to determine if they are required for protein function and tested the ability of these mutated proteins to interact in yeast with known partners. We found that these motifs appear to play at best a minor role in protein function and dimerization capability, despite being strongly conserved. Our results suggest that the functional differentiation of these two paralogous key transcriptional regulators involves both differences in regulation and in sequence; however, sequence changes in the form of unique conserved motifs do not explain the differences observed.

Keywords: APETALA1; FRUITFULL; MADS box genes; conserved protein motifs; functional divergence; gene duplication.

FIGURE 1

Promoter swap and mutated motif constructs. In each construct, the left-hand of the two bars represents the promoter (color-coded, purple for AP1 and blue for FUL) and the right hand bar represents the coding sequence (similarly color coded) Descriptions of the constructs are in the text.

FIGURE 2

FUL can partially substitute for AP1. Flowers from wild type (WT) (A), pAP1:FUL ap1 (B), and ap1 mutants (C). Box plots showing the number of petals per flower (D) and number of flowers per pedicel (E) in WT, pAP1:FUL ap1 lines, and ap1 mutants. Lowercase letters in box plots denote significance; boxes with the same letter are not significantly different from each other according to ANOVAs and Tukey’s Honest Significant Difference tests and following Bonferroni corrections. Cumulative bar graphs describing the identity of first whorl adaxial and abaxial organs (F), first whorl lateral organs (G), and second whorl organs (H) in WT, pAP1:FUL ap1 lines, and ap1 mutants. Sample size is noted below each line. The top number is the number of flowers scored, and the bottom number is the number of plants from which these flowers came.

FIGURE 3

AP1 can partially substitute for FUL. Siliques of WT (A), pFUL:AP1 ful (B), and ful mutants (C). Cauline leaves of WT (D), pFUL:AP1 ful (E), and ful mutants (F). Terminal flower phenotype of pFUL:AP1 ful lines (G). Box plots showing silique length in millimeters (H) and cauline leaf width:length (W:L) ratio (I) for WT, pFUL:AP1 ful lines, and ful mutants. Lowercase letters in box plots denote significance; boxes with the same letter are not significantly different from each other according to ANOVAs and Tukey’s Honest Significant Difference tests and following Bonferroni corrections. Sample size is noted below each line. The top number is the number of siliques or cauline leaves scored, and the bottom number is the number of plants from which they came.

FIGURE 4

AP1 farnesylation motif is not necessary for protein function. Flowers from WT (A), pAP1:mAP1 ap1 (B), and ap1 mutants (C). Box plots showing the number of petals per flower (D) and the number of flowers per pedicel (E) for WT, pAP1:mAP1 ap1, and ap1 mutants. Lowercase letters in box plots denote significance; boxes with the same letter are not significantly different from each other according to ANOVAs and Tukey’s Honest Significant Difference tests and following Bonferroni corrections. Cumulative bar graphs describing the identity of first whorl adaxial and abaxial organs (F), first whorl lateral organs (G), and second whorl organs (H) in WT, pAP1:mAP1 ap1 lines, and ap1 mutants. Sample size is noted below each line. The top number is the number of flowers scored, and the bottom number is the number of plants from which these flowers came.

FIGURE 5

FUL-like motif plays a minor role in silique elongation. Siliques from WT (A), pFUL:tFUL ful (B), pFUL:mFULp ful (C), pFUL:mFULw ful (D), and ful mutants (E). Photographs of cauline leaves from WT (F), pFUL:tFUL ful (G), pFUL:mFULp ful (H), pFUL:mFULw ful (I), and ful mutants (J). Box plots showing silique length in millimeters (K) and cauline leaf width:length (W:L) ratio (L) for WT, pFUL:tFUL ful, pFUL:mFULp ful, pFUL:mFULw ful, and ful mutants. Lowercase letters in box plots denote significance; boxes with the same letter are not significantly different from each other according to ANOVAs and Tukey’s Honest Significant Difference tests and following Bonferroni corrections. Sample size is noted below each line. The top number is the number of siliques or cauline leaves scored, and the bottom number is the number of plants from which they came.

FIGURE 6

The FUL-like, but not the farnesylation, motif may play a role in mediating protein interactions. Interactions in yeast on -HWL 20 mM 3AT (A) and -HWL 30 mM 3AT (B) plates after 6 days of growth. Mutated proteins were compared with WT AP1 and FUL proteins to determine whether altering or abolishing the conserved motifs disrupted protein–protein interactions. Numbers above the images refer to proteins listed in the key on the right hand side. Constructs fused to the binding domain are labeled “DB”; constructs fused to the activation domain are labeled “AD.” Proteins and controls are described in the text. + and – indicate positive and negative (empty vector) control. AP1-PGA, mAP1-PGA = AP1 protein and AP1 protein with mutated farnesylation motif lacking the proline-rich (P), glutamine-rich (G), and activation (A) domains to abolish autoactivation (see Supplementary Figure S2A for diagram).