Stability-activity tradeoffs constrain the adaptive evolution of RubisCO

Abstract

A well-known case of evolutionary adaptation is that of ribulose-1,5-bisphosphate carboxylase (RubisCO), the enzyme responsible for fixation of CO2 during photosynthesis. Although the majority of plants use the ancestral C3 photosynthetic pathway, many flowering plants have evolved a derived pathway named C4 photosynthesis. The latter concentrates CO2, and C4 RubisCOs consequently have lower specificity for, and faster turnover of, CO2. The C4 forms result from convergent evolution in multiple clades, with substitutions at a small number of sites under positive selection. To understand the physical constraints on these evolutionary changes, we reconstructed in silico ancestral sequences and 3D structures of RubisCO from a large group of related C3 and C4 species. We were able to precisely track their past evolutionary trajectories, identify mutations on each branch of the phylogeny, and evaluate their stability effect. We show that RubisCO evolution has been constrained by stability-activity tradeoffs similar in character to those previously identified in laboratory-based experiments. The C4 properties require a subset of several ancestral destabilizing mutations, which from their location in the structure are inferred to mainly be involved in enhancing conformational flexibility of the open-closed transition in the catalytic cycle. These mutations are near, but not in, the active site or at intersubunit interfaces. The C3 to C4 transition is preceded by a sustained period in which stability of the enzyme is increased, creating the capacity to accept the functionally necessary destabilizing mutations, and is immediately followed by compensatory mutations that restore global stability.

Fig. 1.

The RubisCO hexadecamer structure. Pairs of large subunits (blue and yellow) form dimers with an extensive interface; four of these dimers form an octomeric ring. The interdimer interfaces are comparatively small, and the overall structure is stabilized by the binding of eight small subunits (lavender) that bridge dimers. (A and B) Surface views from side and top, respectively. (C) The two chains forming the L_AL_B dimer are shown in ribbon form. Each dimer forms two active sites, the upper site here being between the N-terminal domain of L_A and the C-terminal domain of L_B. Each site undergoes an open to closed structural transition on substrate binding. The reaction intermediate analog 2-carboxyarabinitol-1,5-bisphosphate is shown bound at each site in this structure (PDB code: 1WDD). The larger C-terminal domain contributes most residues to each active site, but the N-terminal domain is critical for positioning the CO₂ or O₂ molecule. (D) Atoms of residues under positive selection in the large subunit (L_B) are shown as spheres. These residues are frequently close to subunit interfaces.

Fig. 2.

Effect of mutations on protein stability. (A) Stability landscape of the large subunit (rbcL). All 19 possible mutations at each position observed in the O. sativa structure (positions 12–456) are colored on a vertical bar in terms of their stability relative to the native residue. Residues that are part of the active site are indicated by a black bar. The thresholds for ΔΔG_fold in kcal/mol are highly stabilizing (< −1.84), stabilizing (−1.84 to −0.92), slightly stabilizing (−0.92 to −0.46), neutral (−0.46 to +0.46), slightly destabilizing (+0.46 to +0.92), destabilizing (+0.92 to +1.84), and highly destabilizing (> +1.84). Positions where the vertical bar is substantially gray or blue are predicted to be tolerant of mutation and where largely red are intolerant. Highly destabilizing mutations are very unlikely to occur in nature. (B) Stability effect of observed mutations at each position, relative to the O. sativa rbcL sequence. Within the monocot species, 105 positions of the 444 aligned residues of the peptide chain have alternate amino acids. The overwhelming majority of observed mutations (79.5%) have modest stability changes in the range of −1.84 to +1.84 kcal/mol.

Fig. 3.

Distribution of stability effects of possible mutations and those occurring during evolution. The distribution of stability changes arising from mutations observed in the evolutionary history of the reconstructed ancestral sequences (solid line) stands in contrast to that of all possible simulated mutations (dashed line). Both distributions have their largest peak close to a ΔΔG of zero. The observed mutations have an excess of slightly stabilizing observed mutations and also a distinct peak of slightly destabilizing and destabilizing values centered at +0.88 kcal/mol. The majority of possible mutations are highly destabilizing and rarely occur during evolution. The probability distributions shown here are obtained by kernel smoothing of the original data (Fig. S1).

Fig. 4.

Stability effect and location of ancestral mutations. The 751 mutations occurring during evolution are separated in A by their selection constraints: negative selection or neutral evolution (P > 0.05 from TDG09 after false discovery rate correction), positive selection (0.01 < P < 0.05), and strong evidence of positive selection (P < 0.01) and binned according to their stability effect. (B) Mutations are separated into their branch type (C₃→C₃, C₃→C₄, or C₄→C₄) and binned by their stability effect. (C) Mutations are classified following the subunit interface definitions in ref. and Fig. S2: Intra are in contact only with other residues of the same large subunit, LL1 residues are in contact with the other large subunit of the same dimer (e.g., the L_AL_B interface), LL2 and LL3 contact a large subunit of another dimer (e.g., L_BL_C and L_BL_D, respectively), and LS are all residues in contact only with any of the small subunits. (D) Mutations are separated into their branch type and binned into their contact interfaces. Categories are highlighted by an * when enriched or an “o” when depleted.

Fig. 5.

Changes in stability through evolution. (A) Frequency of mutations in each category of stability against their evolutionary branch positions relative to the C₃→C₄ transition. There is a long period in which slightly stabilizing mutations are accumulated before the transition in which a substantial number of destabilizing and slightly destabilizing mutations occur. In the branch following the transition, there is a peak of apparently compensatory stabilizing or slightly stabilizing mutations. Stability categories as in Fig. 2. (B) Cumulative mean net change in stability in the neighborhood of the C₃→C₄ transition. (C) The corresponding cumulative mean contributions to stability of all stabilizing and all destabilizing mutations (the latter is offset by −5 kcal/mol to aid comparison).