Darwin shines light on the evolution of bioluminescence

Abstract

The presence of luminous organs in a few insects, belong-ing to different families and orders, offers a parallel case ofdifficulty (to the origin of the electric organs of fishes).Thus wrote Charles Darwin in ‘Difficulties on Theory’ in On theOrigin of Species (1). He couldn’t see how small change by smallchange could lead, apparently out of the blue, to a completelynew phenomenon such as the electric organs of fishes or theluminous organs of fireflies and jellyfish through his BIG idea ofnatural selection. A particular puzzle, pointed out by Darwin, isthe apparently random distribution of luminous species within aphylum, or even a genus. This problem was highlighted by E.N.Harvey, who likened the phylogenetic distribution of biolumines-cence to ‘a handful of wet sand being thrown at a blackboardcontaining a list of all genera’ (2), the luminous organisms beingidentifiedbywherethesandstuck.Thedistinguished astrophysistFred Hoyle argued strongly that Darwin’sideawouldnotwork.The mathematics just didn’t add up. Hoyle compared evolutionthrough random mutations to how ‘a tornado sweeping througha junk-yard might assemble a Boeing 747 from the materialstherein’ (3). But, the recent article published in the May –June issueof Luminescence (4) shines light on both Darwin’sandHoyle’sdilemma. Far from being a problem for evolution, this latest papershowsthatbioluminescenceisamodelforoneofthekeyproblemsin evolution – the origin of a new enzyme. All that is needed is asolvent cage, within which are just a few critical amino acids (4).Bioluminescence, or ‘phosphorescence’ as Darwin called it,is the emission of visible light from living organisms (5). Itoccurs in 8 phyla, Protozoa, Cnidaria, Ctenophora, Mollusca,Echinodermata, Arthropoda, Chaetognatha and Chordata,and is a major communication system in the deep sea. Allbioluminescence is chemiluminescence, the generic reactioninvolving the oxidation of a small organic molecule, the luciferin,catalysed by an enzyme, the luciferase, or photoprotein in somecases. All bioluminescence requires oxygen, in some form,because the energy required to excite an electron for theemission of visible light comes from oxidation. This is true evenin fireflies and glow-worms, which use ATP to form the realluciferin, AMP–luciferin. ATP does not have an energy-rich bond,and is not the energy source for firefly or any other biolumines-cence, as is often wrongly stated in several textbooks andsome websites! There are five recognized chemical families ofluciferin causing bioluminescence: flavines (e.g. bacteria);imidazopyrazines (e.g. hydroids, jellyfish, sea pansies and pens,decapod shrimp, some squid and fish);benzothiazole (e.g.firefliesand glow-worms); linear tetrapryolles (e.g. dinoflagellates andeuphausid shrimp); and aldehydes (e.g. earthworms and the snailLatia).However,theremustbeseveralotherchemistriesthathaveyet to be revealed, as these uncharacterized bioluminescent sys-tems do not cross-react with known luciferases. This molecularbiodiversitymeansthateachtypeofchemistrymusthaveevolvedindependentlyfroma differentorigin.Yettheevolutionaryoriginsof none are known.The latest paper relating to the evolution of bioluminescence(4) focuses on the most common chemistry causing biolumines-cence in marine organisms – coelenterazine. This wasdiscovered in the coelenterate jellyfish Aequorea by NobelLaureate Osamu Shimomura (6,7). Ironically, this may be amisnomer, because luminous jellyfish do not appear to be ableto synthesize coelenterazine de novo, and therefore must obtainit through the food chain. Coelenterazine is predicted to besynthesized from the cyclization of three amino acids – Phe,Tyr, Tyr – which is not a problem via random mutation, refutingHoyle’s argument (3). The most credible evolutionary origin forcoelenterazine is as an oxygen metabolite scavenger analogousto vitamin C in our blood. But the puzzle for the molecularbiologist, highlighted in this latest paper (4), is that when theknown amino acid sequences for coelenterazine luciferasesand photoproteins are compared there is very little similaritybetween them, even within quite closely related phylogeneticgroups. So how could such diversity in protein sequencescatalyse the same chemical reaction to produce light?Vassel et al. used albumin as a model for a primeval luciferase(4), because it is known to bind a wide range of substances,and was predicted to form a solvent cage that could providethe necessary electrochemical environment to catalysecoelenterazine chemiluminescence. The results show, for thefirst time, that albumin does indeed catalyse coelenterazinechemiluminescence, constistent with a mono-oxygenaseenzymatic activity. Evidence for this being truly enzymatic wasbased on the fact that heat denaturation destroyed the activity,it was saturable by the substrate coelenterazine, it increasedwith alkaline pH, and it was blocked by several divalent cationsand by drugs known to bind to albumin. Using modellingsoftware, the coelenterazine binding site was identified asSudlow site 1. Only drugs that bound this site inhibited the