Biosensing Firefly Luciferin Synthesis in Bacteria Reveals a Cysteine-Dependent Quinone Detoxification Route in Coleoptera

Note on attribution: this is from the Viviani lab, Vadim Viviani is the senior/corresponding author, but de Souza is first author. The paper is sometimes cited as “Viviani et al.” for that reason, but the correct first-author citation is de Souza et al.

• A bioluminescent biosensor for in-cell luciferin synthesis. E. coli BL21 immobilized in agar wells and transformed with the bright Amydetes vivianii firefly luciferase (AmyLuc) emit light only if D-luciferin is present in the cytoplasm, making bioluminescence itself the readout for whether luciferin is being synthesized inside the cell. Add cysteine and p-benzoquinone to the well; if light comes on, luciferin formed; if it doesn't, either no luciferin formed or the cells are dead. Cell viability is then verified separately by adding exogenous commercial D-luciferin. Same setup also worked in Pichia pastoris with Pyrearinus termitilluminans click beetle luciferase. This is a clean, fast, scalable assay format that any synthetic-biology lab could adapt for screening luciferin-precursor combinations or quinone-detoxification candidates.
• D-cysteine + benzoquinone produces D-luciferin instantly; L-cysteine + benzoquinone produces it slowly via in-cell racemization. D-cys + BQ in either buffer or transformed E. coli gave intense, fast bioluminescence, direct confirmation in cells of the spontaneous chemistry Kanie 2016 demonstrated in buffer. L-cys + BQ gave weaker, delayed bioluminescence (peaking around 30 min), consistent with formation of L-luciferin followed by racemization to D-luciferin via the luciferase/ATP/CoA/esterase pathway Niwa 2006 outlined. Both enantiomers of cysteine are competent precursors; the difference is just kinetics and the requirement for endogenous racemization machinery (CoA, esterases, present in E. coli, also present in plant cells). Hydroquinone + cysteine also produces luciferin via spontaneous oxidation of HQ to BQ at physiological pH; dopamine + cysteine also works, though less efficiently, presumably via dopamine-quinone.
• Direct experimental confirmation that laccase enhances luciferin biosynthesis from hydroquinone. Commercial Rhus vernicifera laccase incubated with hydroquinone and D-cysteine at pH 6.5 produced substantially more luciferin (measured by AmyLuc bioluminescence) than the no-laccase control. The bioluminescence emission spectrum of the laccase-derived product with AmyLuc peaked at 551 nm, within 4 nm of the spectrum from authentic commercial D-luciferin (547 nm), confirming the product is genuinely D-luciferin. This is the cleanest direct enzymatic precedent in the literature for laccase as a productive component of a heterologous luciferin biosynthesis pathway, and the strongest single piece of evidence supporting AtLAC17 as a TU5 gene choice.
• The benzoquinone-toxicity / cysteine-rescue result is the conceptual core of the paper. p-benzoquinone has IC50 ~25 µM against E. coli (~100 µM in P. pastoris), quinones are extremely cytotoxic because they react with thiols and amines in proteins and generate reactive oxygen species. But in the presence of cysteine, bacteria remained viable at 3.6 mM BQ, a 144-fold rescue, with concomitant luciferin formation. The interpretation is that free cysteine in the cytoplasm out-competes protein thiols and amines for the reactive quinone, forming the cysteinyl-hydroquinone Michael adduct (Kanie 2018) and ultimately luciferin, sparing the proteome. Cells under quinone stress effectively want to do this reaction, because the alternative is enzyme inactivation and death.
• A new evolutionary thesis: luciferin originated as a detoxification byproduct of failed sclerotization. The authors re-analyzed transcriptomes of firefly lanterns (Lampyridae), railroad worms (Phengodidae), click beetle photogenic tissues (Elateridae), and non-bioluminescent control beetles. Lanterns express the upstream sclerotization/tanning enzymes, tyrosine hydroxylase, DOPA decarboxylase, laccase 2, dopamine N-acetyltransferase, but specifically lack the terminal-step enzymes dopachrome isomerase (DCE) and NBAD synthetase. Without these terminal enzymes, dopamine and dopamine-quinone accumulate to potentially cytotoxic levels. Lanterns simultaneously upregulate cysteine biosynthesis (cystathionine-β-synthase, cystathionine-β-lyase). High cysteine + accumulated quinones → spontaneous luciferin formation as a detoxification byproduct. The evolutionary argument predicts (correctly) that bioluminescence should evolve in soft-bodied / less-sclerotized beetle lineages, Lampyridae, Phengodidae, Rhagophthalmidae are all soft-cuticled, and the hard-bodied Elateridae are bioluminescent mainly in larvae and in less-sclerotized cuticular regions of adults. Whether this thesis is correct or not, it is the cleanest “why fireflies?” story currently in the literature.
• The cellular context for the spontaneous chemistry is now experimentally established. Kanie 2016 demonstrated BQ + cys → luciferin in buffer; Oba 2013 demonstrated it in vivo in adult firefly lanterns; Kanie 2018 identified the cysteinyl-HQ intermediate in pupae. de Souza 2022 closes the loop by showing the same chemistry runs in heterologous cellular contexts (E. coli, P. pastoris) just by adding the substrates. The reaction does not require firefly enzymes, firefly cell types, firefly compartmentation, or firefly evolutionary history, it is a default outcome whenever cysteine and a quinone meet in a cell. This collapses several potential mechanistic uncertainties about heterologous reconstitution.

Bottom line for the project: This paper is direct, in-cell, experimental validation of the strategic premise underlying the pivot away from a fully heterologous luciferin pathway and toward exploiting endogenous plant phenolic chemistry. The case for the four-TU minimum design (TU1 luc2+SKL, TU2 PPYR_02911, TU3 BGLU46+SKL, TU4 ACOT9, TU5 AtLAC17) becomes substantially stronger with this citation in the bibliography for three reasons. First, the laccase result is the strongest enzymatic precedent in the literature for TU5 (AtLAC17), laccase oxidation of hydroquinone to benzoquinone is a confirmed productive step in luciferin biosynthesis, not just a guess from co-expression data. Second, the cysteine-rescue toxicity finding means N. tabacum leaves should tolerate high benzoquinone concentrations as long as endogenous cysteine pools can keep up, and tobacco runs active cysteine biosynthesis through its sulfur assimilation pathway, so the cysteine side of the reaction is unlikely to be limiting. Third, the spontaneous-chemistry-in-cells finding means there is no need to engineer a full enzymatic luciferin biosynthesis pathway; the BQ + cys → luciferin reaction will happen on its own once both substrates are in the same compartment, which is exactly the assumption the project's strategic pivot rests on. Beyond the design justification: the bioluminescent-biosensor methodology is also worth knowing as a potential QC assay for the project, express luc2 in E. coli, feed candidate precursor combinations, watch for light. This would be a fast iteration tool for testing whether candidate enzymes from different organisms produce a luciferin-like output before committing to MoClo construction. Cite de Souza 2022 alongside Kanie 2016, Kanie 2018, and Oba 2013 as the four-paper biochemical foundation that the construct design rests on, with Zhang 2020 supplying the ACOT half of the story.