Skip to content
Skip to content

Independent e-magazine

the OUTSPOKEN digest

MIT Wired Bacteria Into Transistors. Its Biggest Circuit Is Twenty-Four Colonies on a Dish That Takes Eight Hours to Add

Engineered colonies of a plant bacterium pass chemical signals across five millimetres of agar, switching each other on and off like components on a board. The team built adders, routers and logic gates. The point is a computer that can grow on a leaf.

Outspoken Digest Technology Desk

Tuesday, September 8, 2026/3 min read

Bacterial colonies grown in patterns on an agar plate. These are E. coli; the MIT circuits use colonies of Pantoea agglomerans printed onto agar in the same way
Photo: LadaSy via Wikimedia Commons (CC BY 4.0)

A transistor is a switch that one signal controls and another passes through. Engineers at MIT have built one out of a bacterium. Then they built two dozen of them on a single Petri dish, wired them together with three more strains acting as relays, and used the result to add numbers, route signals to chosen destinations and perform logic. The work, led by Hamid Doosthosseini in Christopher Voigt's laboratory in MIT's Department of Biological Engineering, was published in Nature Chemical Biology on 17 August.

How a colony becomes a component

The bacterium is Pantoea agglomerans, an unremarkable organism that lives on the surfaces of plants. The team engineered two versions of it that behave as transistors. Each responds to a control molecule called OC-6, which switches it on or off, and when on, each detects a second molecule, OC-12, and in response produces a third, OHC-14, as its output. Three further strains act as relays, translating one signal into another so that the output of one transistor can become the input of the next.

The circuit is laid out in space. Colonies are printed onto agar about five millimetres apart, and the signalling molecules diffuse from one to the next. Wiring, in this system, is geography: put a relay between two transistors and they are connected, leave a gap and they are not. From those parts the team assembled multi-input logic gates, an OR gate, an imply gate, a two-input adder, a demultiplexer that routes a signal to a chosen output, and a multi-signal processor. The largest network had twenty-four colonies linked by chemical communication.

The catch, which is also the point

A calculation takes about eight hours, the time for the molecules to diffuse and the cells to respond. Nobody is going to run software on this. The comparison with silicon is the wrong frame, and the researchers do not make it.

The right frame is where the computer can be. Pantoea already lives on leaves and roots. A circuit made of it could, in principle, coat a plant, sense the chemical signature of drought or a pest attack, process that information across a few colonies, and trigger a response such as synthesising a fungicide, all without a battery, a chip or a person. None of that has been demonstrated yet; the paper stops at the Petri dish. But the reason to build a slow computer out of bacteria is to put computation in places no other computer can go, and the reason to build it as transistors rather than as a single engineered cell is that transistors compose. A designer can add a stage without redesigning the whole.

Limits

The team is candid about two. The supply of transcription factors, the molecular parts that let a cell respond to a specific signal, is limited, which constrains how many distinct signals a circuit can use. And packing complex circuitry into a single cell burdens it with protein production until it stops growing properly, which is precisely why spreading the logic across many colonies, each doing one simple thing, is the more scalable route. The work was funded by DARPA and IARPA, the US defence and intelligence research agencies, which tells you who expects to want sensors that grow.

This is the second piece of synthetic biology in a week to move a boundary that looked fixed. The eight-letter DNA result showed a natural enzyme reading an artificial genetic alphabet; this one shows natural cells doing arithmetic. Neither is a product. Both are the sort of result that, a decade later, turns out to have been the beginning of one. For the everyday biology of the bacteria that already live on us and in us, and what can honestly be said about manipulating them, see our piece on microbiome testing.

Published in The Outspoken Digest

Editorial desk

Outspoken Digest Technology Desk

Software, hardware, artificial intelligence and what they change for everyone else.

Newsletter

The Digest, in your inbox

One edition, sent when it is ready. No noise, and your address is never passed on.

We send a confirmation first. One click to leave, always.

Share this story

the OUTSPOKEN digest

Beyond boundaries. Independent stories on technology, culture, and the trends shaping how we live.