Plants with Genetically Encoded Autoluminescence

• First autonomously bioluminescent plants visible to the naked eye, no exogenous substrate required. Tobacco (N. tabacum and N. benthamiana) lines carrying the four-gene fungal cassette, nnLuz, nnHispS, nnH3H, nnCPH, codon-optimized for tobacco and integrated via Agrobacterium-mediated random nuclear insertion. Light output is roughly an order of magnitude brighter than the previous bacterial-lux autoluminescent plant attempts (Krichevsky 2010), and visible to consumer cameras with 0.5 to 30 second exposures. This is the proof of concept that anchors the entire commercial bioluminescent-plant space and is the direct precursor to Light Bio's Firefly Petunia.
• Caffeic acid as endogenous precursor, the pathway plugs into native phenylpropanoid metabolism. Plants make caffeic acid as a normal intermediate in the phenylalanine → coumarate → caffeate → lignin/flavonoid axis, so the fungal four-enzyme cassette taps an existing flux pool without needing imported precursor genes. Notably, no NpgA (phosphopantetheinyl transferase) was required for HispS activity in tobacco, in contrast to P. pastoris (Kotlobay 2018), suggesting plant native PPTases activate the polyketide synthase. Substrate-feeding experiments showed that exogenous caffeic acid increased luminescence (slowly), hispidin produced fast bright luminescence, and ATP/CoA/malonyl-CoA did not, pinning caffeic acid as the rate-limiting precursor in planta.
• No detectable growth or developmental burden. 15 independent transgenic lines were assayed. Chlorophyll, carotenoid content, flowering time, seed germination, and overall morphology were indistinguishable from wild-type, with only a 12% median height increase in transgenics. This is a striking contrast to the bacterial lux system, which is toxic in eukaryotes (n-decyl aldehyde from LuxAB is broadly cytotoxic). The fungal system's tolerance is one of its biggest practical advantages.
• Spatial and temporal luminescence patterns mirror phenylpropanoid flux. Bright at root tips, root branching points (hours before lateral roots are morphologically visible), terminal/axillary buds, young shoots, and especially flowers; dim in older leaves. Light emission tracks endogenous caffeic acid availability, which makes the system not just a static reporter but a dynamic one, the plants light up in response to wounding, methyl jasmonate treatment, and ethylene exposure (banana skin), and luminescence spreads from injury sites along veins at ~2 µm/s. This is unintended but biologically interesting: an autoluminescent plant is a real-time phenylpropanoid-flux reporter.
• The luminescence is continuous, dim, and green-shifted (~520 nm), not flash-bright. Bright enough for the naked eye and consumer cameras, but the photometric ceiling is set by endogenous caffeic acid pool size and by HispS turnover. Flowers reached ~10¹⁰ photons/min; this is many orders of magnitude below what firefly luc + saturating exogenous D-luciferin produces in the same plant via agroinfiltration. Continuous-glow mode is qualitatively different from the pulsed-flash mode that firefly luciferase enables under physiological control.

Bottom line for the project: This paper is the elephant in the room for any bioluminescent-plant pitch and I should engage with it directly rather than around it. Three honest framings work depending on context. (1) Different photonics, different applications: fungal autoluminescence gives a continuous dim glow that tracks caffeic acid flux, beautiful, commercially proven, and exactly what Light Bio sells; firefly luciferase gives ATP-dependent, peroxisomal, kinetically controllable light that is a much better substrate for sensors, reporters, and any application where you want signal modulation rather than constitutive output. The two systems are not competing for the same product. (2) Substrate provenance and pathway novelty: the fungal system parasitizes native phenylpropanoid metabolism, which is elegant but constrains the plant to whatever flux it already had; the firefly system requires building a heterologous luciferin pathway from scratch, which is harder but means the photometric ceiling is set by what you engineer rather than by lignin biosynthesis. (3) Brightness ceiling: the firefly system, when fully autonomous, has a higher theoretical brightness ceiling because D-luciferin + ATP + O₂ is a much higher-energy reaction per photon than 3-hydroxyhispidin oxidation, and because ATP is not flux-limiting in a healthy plant the way caffeic acid is. None of this means the project is misdirected, it means the project is solving a different problem than Light Bio solved, and the comparison should be the start of the pitch, not the awkward question at the end. Cite Mitiouchkina 2020 alongside Kotlobay 2018 wherever you contextualize what's been done already in autoluminescent plants.