Genomic and Experimental Data Provide New Insights Into Luciferin Biosynthesis and Bioluminescence Evolution in Fireflies

• Two more firefly genomes, Lamprigera yunnana and Abscondita terminalis. PacBio long-read assemblies of a glow species (L. yunnana, 1,053 Mb) and a flash species (A. terminalis, 501 Mb), bringing the total of published firefly genomes to five at time of publication. Useful as additional comparative anchors but the genomes themselves are not the headline result for the project, the biochemical claims are.
• Direct experimental validation that ACOT converts L-luciferin to D-luciferin. This is the paper's most important contribution from a pathway-completion standpoint. The authors expressed three A. terminalis acyl-CoA thioesterase candidates (AteACOT1, AteACOT4, AteACOT9) in E. coli and tested them in vitro alongside firefly luciferase with L-luciferin, ATP/Mg²⁺, and CoA-SH; only AteACOT1, the cluster-I ortholog with high transcript and protein abundance in lanterns, efficiently produced D-luciferin, monitored on a chiral HPLC column. This is the first direct enzymatic confirmation in an insect ACOT of the deracemization step that Niwa 2006 and Maeda 2017 had outlined with a bacterial TESB stand-in. AteACOT1 is the orthologue of the gene I have cloned as TU4 (PpyrACOT9 in the iGEM/Fallon nomenclature lineage), and the cluster-I clade in the Zhang phylogeny is the same ancestral mitochondrial-ACOT lineage (HomoACOT9/MusACOT9-10 sisters) that Fallon 2018 had identified.
• A revised, more complete luciferin biosynthesis pathway is proposed, but partly speculative. The authors integrate genomics, lantern transcriptomes/proteomes, and Abd-B knockout mutants (CRISPR/Cas9, larvae lacking lanterns) to argue for a route in which (i) cysteine for luciferin comes from anabolism (cystathionine-γ-lyase highly expressed; CDO/CSAD low), (ii) the benzoquinone precursor may derive from tyrosine catabolism via homogentisate → benzoquinone-acetic acid → activation by 4CL → β-oxidation-style acetyl removal by sterol-carrier-protein-X (SCPx) thiolase to give p-benzoquinone, in addition to or instead of the arbutin-stored hydroquinone route. The cysteine-anabolism point is well-supported. The tyrosine to BQ route is interesting but speculative, it rests on co-expression rather than isotope tracing or in vitro reconstitution, and the SCPx thiolase step doing a non-canonical decarboxylative cleavage to yield the BQ backbone is a mechanistic stretch that has not been directly demonstrated. Worth knowing about; not worth building TUs around.
• Peroxisomal targeting of ACOT remains unresolved. AteACOT1 has no detectable PTS1, mirroring the situation in Drosophila and other insect ACOTs. The authors do the deracemization assay in vitro, so the cell-biology question of how a PTS1-less ACOT gets into the peroxisome where luciferase is acting is left open. Practically: for TU4, I have reasonably retained ACOT9 without an SKL tag based on the M. musculus mitochondrial-targeting precedent, but the question of whether the active firefly enzyme actually reaches the peroxisome in a heterologous system is genuinely unresolved across the field, not a project-specific gap, a community-wide one.
• Convergent evolution of beetle bioluminescence is reconfirmed with broader sampling. Synteny analysis across five firefly genomes plus I. luminosus shows that the Lampyridae luciferase locus is conserved among fireflies but not syntenic with the Elateridae luciferase locus, and that the elaterid luciferase clade is sister to a separate Elateridae-specific PACS clade. Estimated divergence of the Lampyridae luciferase ancestor at ~205 Mya predates the Lampyridae/Elateridae split (~174 to 115 Mya); the elaterid luciferase neofunctionalized later (~131 Mya). This corroborates Fallon 2018's parallel-origin conclusion with more taxa. Does not change my design but worth citing alongside Fallon 2018 wherever convergence comes up.
• A taxonomic reshuffling worth noting in passing. Phylogenomics places Lamprigera inside Luciolinae rather than Lampyrinae, contradicting earlier morphology-based classifications. Of marginal relevance to the construct work, but if I ever cite Lamprigera sequences (e.g. for a luciferase or BGL ortholog comparison) the subfamily attribution should follow Zhang 2020.

Bottom line for the project: This paper is the citation for TU4. My ACOT9 choice, a mammalian (Mus musculus) cluster-I ACOT, no SKL, with a mitochondrial targeting sequence, sits in the same phylogenetic clade that Zhang 2020 directly demonstrated to perform L → D luciferin conversion in vitro for A. terminalis AteACOT1. That is the strongest functional precedent in the literature for the role TU4 is meant to play, and it strengthens the case in any proposal or paper that the four-TU minimum-pathway design (TU1 luc2+SKL, TU2 PPYR_02911, TU3 BGLU46+SKL, TU4 ACOT9) covers all the chirality, oxidation, and substrate-liberation steps required for autonomous bioluminescence given an endogenous benzoquinone-or-hydroquinone source. The tyrosine to BQ pathway speculation is intriguing and worth keeping on a watch list, it would be a second potential precursor source if the arbutin/HQ pool in N. tabacum turns out to be limiting in Phase 2, but it is not yet demonstrated chemistry and should not drive construct design.