Beetle luciferases produce different bioluminescence colors from green to red using the same d-luciferin substrate. Despite many studies of the mechanisms and structural determinants of bioluminescence colors with firefly luciferases, the identity of the emitters and the specific active site interactions responsible for bioluminescence color modulation remain elusive. To address these questions, we analyzed the bioluminescence spectra with 6'-amino-D-luciferin (aminoluciferin) and its 5,5-dimethyl analogue using a set of recombinant beetle luciferases that naturally elicit different colors and different pH sensitivities (pH-sensitive, Amydetes vivianii λmax=538 nm, Macrolampis sp2 λmax=564 nm; pH-insensitive, Phrixotrix hirtus λmax=623 nm, Phrixotrix vivianii λmax=546 nm, and Pyrearinus termitilluminans λmax=534 nm), a luciferase-like enzyme (Tenebrionidae, Zophobas morio λmax=613 nm), and mutants of C311 (S314). The green-yellow-emitting luciferases display red-shifted bioluminescence spectra with aminoluciferin in relation to those with D-luciferin, whereas the red-emitting luciferases displayed blue-shifted spectra. Bioluminescence spectra with 5,5-dimethylaminoluciferin, in which enolization is blocked, were almost identical to those of aminoluciferin. Fluorescence probing using 2-(4-toluidino)naphthalene-6-sulfonate and inference with aminoluciferin confirm that the luciferin binding site of the red-shifted luciferases is more polar than in the case of the green-yellow-emitting luciferases. Altogether, the results show that the keto form of excited oxyluciferin is the emitter in beetle bioluminescence and that bioluminescence colors are essentially modulated by interactions of the 6'-hydroxy group of oxyluciferin and basic moieties under the influence of the microenvironment polarity of the active site: a strong interaction between a base moiety and oxyluciferin phenol in a hydrophobic microenvironment promotes green-yellow emission, whereas a more polar environment weakens such interaction promoting red shifts. In pH-sensitive luciferases, a pH-mediated switch from a closed hydrophobic conformation to a more open polar conformation promotes the typical red shift.

Beetle luciferases emit different bioluminescence colors from green to red; however, no clear relationship between the identity of the luciferin binding site residues and bioluminescence colors was found in different luciferases, and it is unclear whether critical interactions affecting emission spectra occur on the thiazolyl or on the benzothiazolyl sides of the luciferin binding site. Through homology modeling and site-directed mutagenesis using our multicolor set of beetle luciferases (Pyrearinus termitilluminans larval click beetle, Pte, λ(max) = 534 nm; Phrixothrix hirtus railroad worm red emitting, PxRE, λ(max) = 623 nm; and Macrolampis sp2 firefly, Mac, λ(max) = 564 nm), we show that the residues C/T311 (S314) play an important role in bioluminescence color determination. Modeling studies indicate that the main-chain carbonyls of these residues are close to both oxyluciferin phenolate and AMP, whereas the side chains pack against second-shell residues. The C311(S314)A mutation considerably red shifts the spectra of the green-yellow-emitting luciferases (Pte λ(max) = 534 to 590 nm; Mac λ(max) = 564 to 583/613 nm) and affects the K(M) values for luciferin and ATP, but not the spectrum of the red-emitting luciferase. On the other hand, whereas the exchange between C/T311 (S314) caused smaller effects on the emission spectra of green-yellow-emitting luciferases, the C311T substitution (naturally found in green-emitting railroad worm luciferases) resulted in the largest reported blue shift in P. hirtus red-emitting luciferase (λ(max) = 623 to 606 nm). Altogether, these results indicate that the stability of residues C/T311 (S314) and the size of the cavity around oxyluciferin phenolate affect bioluminescence colors and suggest, for the first time, the occurrence of a critical interaction between main-chain carbonyls of position 311 (314) residues and oxyluciferin phenolate.

The origin of luciferases and of bioluminescence is enigmatic. In beetles, luciferases seem to have evolved from AMP-CoA-ligases. How the new oxygenase luminogenic function originated from AMP-ligases leading to luciferases is one of the most challenging mysteries of bioluminescence. Comparison of the cloned luciferase-like enzyme from the nonluminescent Zophobas morio mealworm and beetle luciferases showed that the oxygenase activity may have emerged as a stereoselective oxidative drift with d-luciferin, a substrate that cannot be easily thioesterified to CoA as in the case of the l-isomer. While the overall kcat displayed by beetle luciferases is orders of magnitude greater than that of the luciferase-like enzyme, the respective oxidation rates and quantum yields of bioluminescence are roughly similar, suggesting that the rate constant of the AMP-ligase activity exerted on the new d-luciferin substrate in beetle protoluciferases was the main enzymatic property that suffered optimization during the evolution of luciferases. The luciferase-like enzyme and luciferases boost the rate of luciferyl-adenylate chemiluminescent oxidation by factors of 10(6) and 10(7), respectively, as compared to the substrate spontaneous oxidation in buffer. A similar enhancement of luciferyl-adenylate chemiluminescence is provided by nucleophilic aprotic solvents, implying that the peptide bonds in the luciferin binding site of beetle luciferase could provide a similar catalytically favorable environment. These data suggest that the luciferase-like enzyme and other similar AMP-ligases are potential alternative oxygenases. Site-directed mutagenesis studies of the luciferase-like enzyme and the red light-producing luciferase of Phrixotrix hirtus railroadworm confirm here a critical role for T/S345 in luciferase function. Mutations such as I327T/S in the luciferase-like enzyme, which simultaneously increases luciferase activity and promotes blue shifts in the emission spectrum, could have been critical for evolving functional bioluminescence from red-emitting protoluciferases. Through the combination of I327T/S mutations and N-terminal fusion, the luminescence activity of this enzyme was increased to visible levels, with the development of a totally new orange-emitting luciferase. These results open the possibility of engineering luciferase activity in a set of AMP-CoA-ligases.