Functional evolution of an anthocyanin pathway enzyme during a flower color transition

Abstract

Dissecting the genetic basis for the evolution of species differences requires a combination of phylogenetic and molecular genetic perspectives. By mapping the genetic changes and their phenotypic effects onto the phylogeny, it is possible to distinguish changes that may have been directly responsible for a new character state from those that fine tune the transition. Here, we use phylogenetic and functional methods to trace the evolution of substrate specificity in dihydroflavonol-4-reductase (Dfr), an anthocyanin pathway gene known to be involved in the transition from blue to red flowers in Iochroma. Ancestral state reconstruction indicates that three substitutions occurred during the flower color transition, whereas several additional substitutions followed the transition. Comparisons of enzymatic function between ancestral proteins in blue- and red-flowered lineages and proteins from present-day taxa demonstrate that evolution of specificity for red pigment precursors was caused by the first three substitutions, which were fixed by positive selection and which differ from previously documented mutations affecting specificity. Two inferred substitutions subsequent to the initial flower color transition were also adaptive and resulted in an additional increase in specificity for red precursors. Epistatic interactions among both sets of substitutions may have limited the order of substitutions along branches of the phylogeny leading from blue-pigmented ancestors to the present-day red-flowered taxa. These results suggest that the species differences in DFR specificity may arise by a combination of selection on flower color and selection for improved pathway efficiency but that the exact series of genetic changes resulting in the evolution of specificity is likely to be highly contingent on the starting state.

Fig. 1.

Changes in the anthocyanin pathway responsible for the evolution of red flowers in Iochroma. Pathway enzymes (circled) include flavanone 3-hydroxylase (F3H), flavonoid 3’-hydroxylase (F3’H), flavonoid 3’,5’-hydroxylase (F3’5’H), dihydroflavonol reductase (DFR), and anthocyanidin synthase (ANS) (Holton and Cornish 1995). DFR converts dihydroflavonols (DHK, DHQ, and DHM) into leucoanthocyanidins (e.g., leucopelargonidin, LCP), which are subsequently modified to form three colored anthocyanidins (pelargonidin, cyanidin, and delphinidin). These three can be converted into other anthocyanidins (e.g., peonidin and malvidin; not shown). Finally, anthocyanidins are modified by the addition of sugars and other moieties to form anthocyanins. Three pathway changes were involved in the evolutionary transition from blue flowers with delphinidin-derived pigments to red flowers with pelargonidin-derived pigments: the deletion of the F3’5’h gene (indicated with the gray X), the downregulation of F3’h (indicated with the gray arrow), and a shift in DFR substrate specificity (indicated with the bold oval).

Fig. 2.

DFR sequence evolution in Iochroma. (A) Maximum likelihood phylogeny of Dfr. Taxon sampling within Iochrominae includes eight Iochroma species (I.) and two species of Dunalia (D.); two outgroups (Evolvulus and Gentiana) have been pruned. Also, one of the two alleles was randomly pruned from heterozygous individuals (supplementary table S1, Supplementary Material online). Red-flowered species and lineages are in bold. Branches with greater than 95% PP are indicated with an asterisk. (B) Structure of DFR protein, with numbering based on Vitis vinifera (Petit et al. 2007). Sites important for substrate binding based on the V. vinifera crystal structure are indicated with an asterisk along the top. The shaded region (positions 131 − 156) is an area thought to be important for substrate specificity (Johnson et al. 2001). Locations of mutations involved in the evolution of substrate specificity in Iochroma are shown with arrows below. (C) A reduced phylogeny showing the evolution of substrate specificity in the red-flowered lineages. Pie charts indicate the relative activity of each present-day or ancestral protein sequence on the three substrates (DHM, DHQ, and DHK, fig. 1). The most recent common ancestor of the red- and blue-flowered species (Blue Ancestor, BA) and the common ancestor of the red-flowered species (Red Ancestor, RA) are labeled. The five amino acid substitutions between the Blue Ancestor and red-flowered Iochroma gesnerioides lineage are shown with single-letter abbreviations and their position in the sequence (substitutions along other branches of the tree are not listed). The corresponding positions in V. vinifera (Vv) are also noted in parentheses.

Fig. 3.

Relative activities of DFR on the three substrates for different combinations of mutations. 0 indicates mutation absent, 1 indicates mutation present. (A) Relative activity on DHM, DHQ, and DHK for combinations of mutations M1–M3. (0,0,0) corresponds to Blue Ancestor (BA) sequence and (1,1,1) corresponds to the Red Ancestor (RA) sequence. Moving from (0,0,0) to (1,1,1), solid gray lines indicate a significant increase in relative activity on DHK based on the ANOVA results (supplementary table S4, Supplementary Material online). Dashed black lines indicate a significant decrease in DHK activity, whereas dashed gray lines indicate a nonsignificant change in DHK activity. Each pie graph shows the relative activity for a given DFR sequence on each of the three pigment precursors (DHM, DHQ, and DHK). Values are based on least square means from analyses of variance; the value for relative activity on DHK is given inside each pie chart. (B) Relative activity for combinations of mutations M4 and M5. Interpretation of lines, pie graphs, and values as in (A).