Genetically Encodable Bioluminescent System from Fungi

• The complete fungal “caffeic acid cycle” is identified, four enzymes from a ubiquitous plant metabolite to light. The authors map the full pathway in Neonothopanus nambi: caffeic acid → hispidin (by hispidin synthase, HispS) → 3-hydroxyhispidin / fungal luciferin (by hispidin-3-hydroxylase, H3H) → photon + caffeylpyruvate / oxyluciferin (by luciferase, nnLuz) → caffeic acid (by caffeylpyruvate hydrolase, CPH). This is the second-ever fully described luciferin biosynthesis pathway in any organism, after bacterial lux, and the first eukaryotic one. The luciferase nnLuz is a 267-aa protein with no homologs to known enzyme families, a genuinely novel protein family.
• Three genes plus a phosphopantetheinyl transferase make P. pastoris glow on caffeic acid. hisps, h3h, nnluz, and Aspergillus nidulans npgA (which post-translationally activates the polyketide synthase HispS by attaching a phosphopantetheinyl arm) integrated into yeast yields visible bioluminescence in standard media supplemented with caffeic acid. Drop npgA or hisps and the glow fails. This is the minimum heterologous reconstitution.
• Adding three caffeic acid biosynthesis genes makes yeast autonomously bioluminescent on standard media. Tyrosine ammonia lyase from Rhodobacter capsulatus plus the two E. coli 4-hydroxyphenylacetate 3-monooxygenase components feed caffeic acid from endogenous tyrosine. The seven-gene cassette (TAL + HpaB/HpaC + HispS + H3H + nnLuz + NpgA) is the full autonomy package, and is the direct conceptual ancestor of the autoluminescent plants Mitiouchkina et al. would publish in 2020.
• Plants already have caffeic acid as a core phenylpropanoid intermediate. This is the practical reason fungal autoluminescence has translated to plants more readily than firefly luciferin ever has. Tobacco, Arabidopsis, and most plants run high flux through phenylalanine → cinnamic acid → p-coumaric acid → caffeic acid as part of normal lignin and phenylpropanoid metabolism, so heterologous expression of HispS + H3H + nnLuz + CPH (plus NpgA) can in principle tap an endogenous substrate pool without supplemental feeding. In contrast, plants do not natively make D-luciferin or its hydroquinone/cysteine-derived precursors, which is why the firefly route requires importing the entire luciferin biosynthetic chain (TU2 through TU4 in my design) rather than parasitizing existing flux.
• Reporter performance is competitive with firefly luciferase in mammalian systems. Mouse xenografts with CT26 cells expressing nnLuz vs. P. pyralis luciferase, dosed with a mix of fungal and firefly luciferins, give nearly identical IVIS signals. nnLuz is microsomal (predicted N-terminal transmembrane helix), ATP-independent (unlike firefly luc), and the substrate is water-soluble and cell-permeable. These are real advantages for tissue imaging.
• Bioluminescence evolved once in Agaricales fungi via two gene duplications. Phylogenomics across ~24 sequenced Agaricales places the luz duplication at the base of the order, followed by h3h and hisps duplications a few million years later, with retention of nonfunctional or differently-functional paralogs in many nonbioluminescent lineages. The cluster has been lost independently at least six times. This is a clean single-origin convergence story for fungal luminescence, contrasting sharply with the parallel-origin pattern Fallon 2018 documented in beetles.

Bottom line for the project: This paper is the foundational citation for the alternative bioluminescence platform, the one that already has a credible autoluminescent-plant proof-of-concept (Mitiouchkina 2020) and that uses an endogenous plant metabolite (caffeic acid) rather than a synthesized heterologous precursor (D-luciferin). For my project, Kotlobay 2018 is worth knowing well for two reasons. First, in any proposal or pitch, the existence of the fungal route is the obvious “why not just do this?” question, and the answer needs to engage seriously with it: fungal bioluminescence in plants is dimmer per photon than firefly luciferase at peak output, the emission peak is ~520 nm (similar to firefly green) but the kinetics are continuous-glow rather than the bright pulsed flashes that firefly enzymes can produce, and the peroxisomal-targeted firefly system has different photonics for any application that wants flash control or spatial localization. Second, the fungal system is the strongest commercial competitor in the bioluminescent-plant startup space, Light Bio (Planta LLC's spinout) has already commercialized fungal-luminescence tobacco and petunia. The case for continuing the firefly project is not that fungal is bad; it is that firefly luciferase has different optical properties, a much larger biotechnology install base for downstream tooling, and a peroxisomal autonomy story that is qualitatively different from the cytosolic-fungal one. Cite this paper whenever you frame why-firefly-instead-of-fungal in a grant or pitch deck, and engage the comparison head-on rather than ignoring it.