blog

Did evolution of a hydroxynitrile lyase from esterases require a transitional mechanism?

The divergent evolution of esterases to hydroxynitrile lyases ~100 million years ago is puzzling because the two mechanisms are incompatible. Our recent paper suggests that evolution may have used a transitional mechanism for hydroxynitrile cleavage that could continue to catalyze ester hydrolysis.

Approximately 100 million years ago, as dinosaurs roamed the earth, flowering plants and insects diversified and coevolved. Insects fed on flowers and other plant structures, but also benefitted the plants by pollinating them. Plants evolved defenses against insects, such as releasing cyanide. On a molecular scale, this evolution involved changing an esterase enzyme into a hydroxynitrile lyase enzyme. When insects attack plants, these hydroxynitrile lyases (HNLs) release cyanide by catalyzing the reaction below.

The puzzle regarding the evolution of hydroxynitrile lyases (HNLs) from esterases is that the esterase and modern lysine catalyzed HNL mechanisms are incompatible. Evolution of new enzymes is thought to proceed via promiscuous intermediate enzymes. These promiscuous enzymes catalyze the original reaction and also, perhaps less efficiently, the new function. As evolution continues, this promiscuous function improve into enzymes with new function. But if the two mechanisms are incompatible, then how can a promiscuous enzyme exist?

In previous work, we recreated some ancestral enzymes from ~100 million years ago. This recreation involved computer prediction of the sequences of ancestral enzymes based on the sequences of modern enzymes in this family. Next, chemical synthesis of the DNA encoding several ancestral enzymes created the genes. We inserted these genes into bacteria, which then made the enzymes. We isolated and characterized these enzymes.

Figure. A simplified phylogenetic tree of the evolution of hydroxynitrile lyases from esterases. The ancestral esterases (bottom) diversified to create modern esterases and hydroxynitrile lyases (top). Intermediate enzymes (black circles) show increasing abilities to catalyze the HNL reaction (indicated by numbers; number in parentheses indicate the enantioselectivity).

Modern esterases and HNLs shows similarities, but also differences that make them incompatible. Both contain a catalytic triad of serine, histidine and aspartate as well as a hydrogen bonding region called an oxyanion hole. However, esterases contain a glycine (Gly) and methionine (Met) near this region, while HNLs contain a corresponding threonine (Thr) and lysine (Lys). These differences hinder the other reaction.

The ancestral esterase EST2 contained Gly+Met like modern esterases and the ancestral HNL1 contained Thr+Lys like modern HNLs, but an intermediate ancestor, EST3ml contained Asn+Met. Evolution did replace Gly directly with Thr, but first with asparagine (Asn) and then with Thr. Previous work showed that this transitional form can catalyze both ester hydrolysis and hydroxynitrile lyase cleavage, but how is this possible?

One of the modern HNLs (AtHNL) also contains Asn+Met. Our first hypothesis, which we revise below, was that the ancestral enzyme, EST3ml, followed the same mechanism as AtHNL. Previous researchers showed that AtHNL follows an oxyanion mechanism where the oxyanion hole stabilizes the cyanide as it is released from the hydroxnitrile. Thus, we hypothesize that this oxyanion mechanism is a transitional mechanism. In contrast, the modern HbHNL and BmHNL follow a lysine mechanism where the lysine stabilizes the cyanide as it is released from the hydroxnitrile. Another difference between HbHNL and AtHNL is that they favor opposite enantiomers. HbHNL favors the (S)-enantiomer of mandelonitrile, while AtHNL favors the (R)-enantiomer. Did esterases first catalyze the HNL reaction using the oxyanion mechanism, then later switch to the lysine mechanism?

In this latest publication, we attempted to switch modern HbHNL to the AtHNL mechanism by exchanging active site residues. This switch in mechanism should also switch the enantioselectivity. The simple switch of two amino acids from Thr+Lys to Asn+Met in HbHNL yielded inactive enzyme, so we made more substitutions. There are 22 amino acids close the substrate; 6 are identical in both HbHNL and AtHNL, but 16 differ. We changed ten residues guided by similarities to ancestral enzyme EST3ml. This protein, HbHNL-A10H6, catalyzed cleavage of mandelonitrile, but surprisingly did not reverse enantioselectivity. HbHNL-A10H6 lacks Thr and Lys, so it cannot use the Lys mechanism; the lack of reversal in enantioselectivity suggests it also does not use the oxyanion mechanism. Using molecular dynamics modelling, we discovered a third mechanism where Asn stabilizes the leaving cyanide.

Thus, our revised hypothesis is that the last common ancestor of HbHNL (yellow) and AtHNL (light blue), EST3ml (green), may have used an extinct mechanism, the Asn mechanism, to catalyze hydroxynitrile cleavage, different from the known modern mechanisms. This extinct mechanism could have served as a transition from an esterase to the modern HNL mechanisms.

B. J. Jones, Z. Bata, R. J. Kazlauskas (2017) Identical active sites in hydroxynitrile lyases show opposite enantioselectivity and reveal possible ancestral mechanism. ACS Catal. 7, 4221-9; http://doi.org/10.1021/acscatal.7b01108

About the author

Romas Kazlauskas is a professor of biochemistry at the University of Minnesota, Twin Cities where he teaches a course on protein engineering. He is an expert on biocatalysis and engineering enzymes to make them more stable, more selective and able to catalyze new reactions. e-mail: rjk <at> umn.edu