ABSTRACT An enduring question in evolutionary biology is whether novel functions usually evolve before novel structures, or vice versa. This question is pervasive in molecular evolution, as it is implicit in different models of gene duplication, some of which posit function first, others of which posit structure (duplication) first. Here, we show that light catalytic function evolved before ostracod luciferase genes originated by gene duplication. Luciferases catalyze light reactions in bioluminescent cypridinid ostracods. We first used transcriptome sequencing to discover new cypridinid luciferase candidates and closely related genes. We show the candidates catalyze the oxidation of luciferin to produce light, indicating they are luciferases. In addition, we show that genes related to luciferases with the same domain structure also catalyze a light reaction, albeit far less efficiently. These genes, which we name soroluciferase (sluc) and dual-VWD-α (2VWD-α), are unlikely to be luciferases because they are not expressed in the upper lip (the site of bioluminescence). Additionally, 2VWD-α lacks a signal peptide that directs luciferase out of cells to mix with substrate and produce light in the sea. Because luciferase, sluc, and 2VWD-α all catalyze a light reaction, we conclude that this biochemical function preceded the structural origin of luciferases by gene duplication. This function-first result rules out some models of gene duplication that assume new functions follow duplication, and provides a clear case of molecular exaptation, where function preceded structure in an evolutionary novelty.
INTRODUCTION How do complex biological traits such as eyes, flight, metabolic pathways, or flowers, originate during evolution? These biological features often appear so functionally integrated and so complicated, that imagining the evolutionary paths to such complexity is sometimes difficult. As such, some anti-evolution creationists maintain that functionally integrated, complex biological systems are “irreducibly complex”, that they cannot originate by evolution, and therefore they must be the product of “intelligent design”. In fact, structurally and functionally complex systems have originated through evolutionary processes. Here, we use eye evolution as a focus for how to implement a research program to gain understanding of the origin of complex traits. This research program first requires defining the trait in question, followed by inferring the timing of past evolutionary events with comparative methods. By understanding when different components came together, we can begin to understand how they came together, and by making inferences about possible functions and environmental context, we can begin to understand why those components stayed together. Eye evolution is particularly amenable to such a research program because we can use optics to predict function from morphology and we can predict gene functions because much is known about the genetic components of eyes and light sensitivity. This approach leads to a narrative on animal eye evolution that, while still incomplete, is already rich and detailed in many facets. We know that a diversity of sometimes disparate eyes have evolved using functional components that interact with light. Many parts of this light interaction hardware kit are used outside of eyes and were recruited into light receptive organs during evolution. But what were these components doing outside of eyes? Here, we propose the hypothesis that several parts of the light-interacting hardware kit were recruited for use in eyes from ancestral roles related to stress response.