Glowing Plants: From MIT 2017 to China's Claim in 2026

On April 3, 2026, a Chinese startup called Magicpen Bio unveiled a collection of more than twenty glowing plant species — orchids, sunflowers, chrysanthemums — at the Zhongguancun Forum in Beijing. Their representative, Li Renhan, a graduate of China Agricultural University, delivered a line that immediately ran in headlines: «These plants don’t need electricity. They only need water and fertilizer.» International media dubbed the collection «Avatar-like plants.» Reddit blew up — 3,500 upvotes, 200 comments in hours. The Chinese news release is missing only one thing: a peer-reviewed paper. Which makes this the perfect moment to tell the real, much more interesting story of ten years of synthetic biology that stands behind this demonstration.

Round One: MIT Nanobionics, 2017

Illustration. Source: Unsplash

In 2017, Michael Strano’s group at MIT published a paper in Nano Letters that, at the time, looked almost like a trick. They took an ordinary watercress plant, soaked it in a solution containing three types of nanoparticles, and got a plant that glowed in the dark. Not brightly — not bright enough to read under — but visibly, about the level of emergency lighting in a dark corridor.

The mechanism was elegant. The nanoparticles carried three components:

1. Luciferase — the enzyme from fireflies that actually produces light
2. Luciferin — the «fuel» that luciferase consumes in the reaction
3. Coenzyme A — removes reaction byproducts that would otherwise inhibit the process

The nanoparticles delivered these components into the leaf’s interior cells. The result was a kind of «biological lamp» powered by a chemical reaction. The plant didn’t lose any of its normal functions — it just now also emitted light. Glow duration: about four hours. Then the fuel ran out and the plant went dark, like a lamp with a dead battery.

It was an important demonstration. For the first time, a plant had been made to glow without a continuous external energy source. But it was clear this wasn’t the solution for real applications — like street lighting. You can’t soak trees in nanoparticles every evening.

Luciferase and luciferin — a pair of «enzyme and substrate» that in nature powers bioluminescence in fireflies, some bacteria, deep-sea fish, and certain mushrooms. The reaction: luciferase oxidizes luciferin in the presence of oxygen, and part of the energy emerges as a visible-light photon. The reaction’s efficiency is surprisingly high — about 90% of the energy becomes light, with almost no heat loss.

Round Two: The Fungal Route, 2020

The real breakthrough came three years later, and it came not from America but from Russia — with an international team led by Karen Sarkisyan, a synthetic biologist then working at the Institute of Bioorganic Chemistry in Moscow, later at MRC London Institute of Medical Sciences. In 2020, his group published a paper in Nature Biotechnology with a simple but revolutionary title: «Plants with genetically encoded autoluminescence.»

What was the idea? Until that moment, every attempt to produce a bioluminescent plant ran into the same problem: the light-emitting reaction needs a constant supply of luciferin, and plants don’t make it. You either add it externally (like at MIT), or transplant a firefly enzyme with no substrate inside the plant cell to work on.

Sarkisyan and colleagues found an unusual solution: they turned to the fungal bioluminescence system. The thing is, certain luminous mushrooms (like Neonothopanus nambi) use a compound called caffeic acid to synthesize their luciferin — and caffeic acid is abundant in virtually every plant as a byproduct of phenylpropanoid metabolism. In other words, the «starting material» for fungal luciferin is already present in plants. You just need to insert the genes that convert caffeic acid into luciferin and back, closing the metabolic loop.

The team took four genes from Neonothopanus nambi, optimized them for plant expression, and inserted them into tobacco (Nicotiana tabacum — the classic model organism for plant genetic engineering). The result was stunning. The tobacco plant glowed in the dark — brightly, evenly, continuously. All on its own, without external feeding, batteries, or nanoparticles. The glow intensity was about 10 times stronger than all previous attempts combined.

And it wasn’t a one-day effect. The plant kept glowing throughout its life cycle — from seedling to flowering. Brightness varied with cell activity, time of day, and caffeic acid levels. Young, actively growing parts glowed more brightly than mature ones. In effect, you could visually observe the plant’s metabolic activity in real time.

From Lab to Flower Shop

Sarkisyan’s 2020 paper quickly moved beyond academia. One of the authors founded a U.S. startup called Light Bio that licensed the technology and began commercializing it. The company’s first product — the Firefly bioluminescent petunia — was approved by the U.S. Department of Agriculture in 2024 for sale as a regular houseplant.

The Firefly Petunia has been on sale through Light Bio’s website for a couple of years now. It looks like an ordinary white petunia by day, but at night its flowers and buds emit a soft green glow — bright enough to see with the naked eye in a dark room, though not enough to illuminate the room itself. Customers get a plant with documentation, and sales are regulated like any GMO intended for home use.

It’s the first commercially available bioluminescent living organism specifically engineered for sale. For three decades before this, scientists had made glowing jellyfish, glowing fish, and glowing trees — but all were research tools or curiosities. The Firefly Petunia is the first one you can actually buy and put on your windowsill.

Round Three: China, 2026

And this is where Magicpen Bio enters the story. Li Renhan and his team brought a booth to the Zhongguancun Forum displaying globally familiar ornamental species — orchids, sunflowers, chrysanthemums — each glowing in the dark. Striking. Impressive. The videos spread across networks in hours.

But if you read the details, a list of questions piles up. Which specific genes were inserted? The announcement says «from fireflies and luminous fungi, ” which sounds like a combination of the MIT and Sarkisyan pathways. But no concrete genes are named. What’s the brightness in numerical units (candelas per square meter, for example)? No data. How long does one plant glow — an hour, a day, a lifetime? No data. Where is the peer-reviewed paper? No link. Public sources mention only the forum demonstration and the company’s commercial plans.

This doesn’t mean the finding is a fake. It’s entirely possible Magicpen Bio extended the Sarkisyan technology to other species or improved its parameters. Biologically it’s plausible: the fungal pathway that works in tobacco should, in principle, work in other plants with sufficient caffeic acid in their tissues. But without a peer-reviewed publication and without independent validation, we can’t tell extension of the technology apart from a marketing stunt.

The right stance for anyone evaluating the scientific significance is cautious optimism with an asterisk. The direction is real, the idea works, and if Magicpen Bio has genuinely gone beyond Light Bio’s petunias, it would be a good incremental step. But turning this into «the future of city lighting» is premature. At minimum, wait for a peer-reviewed paper with measurements. At maximum, wait for independent replication.

Can a Plant Actually Light a Street?

Short answer: not yet. Let’s do the math.

A typical streetlamp produces about 4,000–8,000 lumens — enough to light a few meters of sidewalk. A Firefly Petunia produces roughly 0.1 lumens per plant. To replace one streetlamp, you’d need about 40,000 to 80,000 petunias. Physically possible — a dense carpet of glowing plants along a sidewalk — but it doesn’t look like what anyone would call city lighting. And it’s still dimmer than the evening spill of lit windows from a city block.

A more realistic niche is decorative lighting — in parks, home interiors, along garden paths. Here, bioluminescent plants could replace low-voltage LED string lights. Environmentally it may even make sense, because plants generate the light from the solar energy they collected during the day — unlike LEDs that need grid electricity.

But even for decorative use, real limitations remain. The emission color is blue-green (peak fungal luciferin emission is around 520 nm). That’s a fairly narrow band; getting a warm yellow or red glow is significantly harder and requires different genetic pathways. The intensity depends on the plant’s health: a healthy plant glows brightly, a stressed or dried one dims. And the duration: «glow» happens only at night, when caffeic acid reserves are high; come sunrise, metabolism shifts and the light fades.

What’s Next

Real breakthroughs in the next 5–10 years will probably come along several lines. First — brightness gains. Multiple labs are working to make plants 10 to 100 times brighter than today’s samples. Second — color palette expansion. Genetic modifications to the luciferase enzyme can shift emission toward red or blue. Third — species scaling. The fungal pathway works well in tobacco, petunias, and apparently in the species Magicpen showed off. But for real application, you need trees and shrubs — whose genetics are more complex and whose life cycles are longer.

And fourth — the regulatory framework. Any genetically modified plant in an open urban environment is a serious regulatory question. The EU, U.S., and China have very different rules on this. The Firefly Petunia was approved in the U.S. precisely because it’s non-invasive, doesn’t produce viable seeds, and isn’t meant for wild release. A street tree is an entirely different story.

The underlying technology for this whole story is published in the peer-reviewed Nature Biotechnology. The 2026 Magicpen Bio demonstration is, as of this writing, not accompanied by a formal scientific publication.

The main thing to take from this story: glowing plants are not fantasy and not hoax — they’re a real working technology. But the mature form isn’t a replacement for street lighting. It’s a beautiful, technically interesting decorative object. And the road to it took ten years, through three generations of approaches: nanobionics in 2017, the fungal pathway in 2020, and extension attempts in 2024–2026. Each step was incremental, and none was an «Avatar-level» revolution. The revolution hides in the small print — because everything in biology that gets done for real gets done in very small steps.